Accounting, Organizations and Society xxx (xxxx) xxx

Contents lists available at ScienceDirect

Accounting, Organizations and Society

journal homepage: www.elsevier.com/locate/aos

Accounting, simultaneity and relative completeness: The sales and

operations planning forecast and the enactment of the ‘demand chain’
Lichen Yu a, *, Jan Mouritsen b
a
Discipline of Accounting, The University of Sydney Business School, The University of Sydney, Australia
b
Department of Operations Management, Copenhagen Business School, Denmark

a r t i c l e i n f o a b s t r a c t

Article history: This study adds to the literature on accounting incompleteness and instability an understanding of how
Received 22 October 2017 accounting acts upon an object that is practised as a multiple. It explores how accounting, in the form of a
Received in revised form sales and operations planning (S&OP) forecast, helps discover the objections attached to the various
27 February 2020
enactments of an object multiple, namely a ‘demand chain’ - a lateral ordering of the firm’s production
Accepted 13 March 2020
Available online xxx
and products from customers backward to suppliers - and translate these objections into decision
mechanisms. The paper finds a process of accumulation of accountings that contributes to the enactment
of the ‘demand chain’ multiple. In this process, accounting becomes both performative and provocative.
As a source of performativity, accounting is relatively complete because it turns each emerging objection
into a specific decision model, enacting the ‘demand chain’ in a certain way. As a force of provocation,
accounting stimulates new objections to emerge against what accounting reveals about the ‘demand
chain’; this adds new accountings onto existing ones, all of which exist simultaneously.
© 2020 Elsevier Ltd. All rights reserved.

1. Introduction production and products from customers backward to suppliers ‒

through a concern about internal versus external customers. The
‘ … factories put sales’ forecast aside as they speak a different demand chain multiple is practised in different ways and raises the
language. Sales always consider the volume sold to customers, question: how may accounting for multiplicity exist? This is a
but the factories always think in terms of shipment from either different question from that asked by studies exploring the com-
warehouses or factories to customers, and shipment from fac- plications of inscription work related to incompleteness (Ahrens &
tories to factories’ (Product Line Planning Manager A e Indus- Chapman, 2004, 2007; Busco & Quattrone, 2015, 2018; Jordan &
trial Division, EuroTech) Messner, 2012; Jørgensen & Messner, 2010; Millo & MacKenzie,
2009) and instability (Andon, Baxter, & Chua, 2007; Chua, 1995;
Dambrin & Robson, 2011; Quattrone & Hopper, 2001; Revellino &
The multiplicity of an object ‒ what Law and Singleton (2005)
Mouritsen, 2015). Those streams of research focus on managers’
and Mol (2002) refer to as an object multiple ‒ adds complications
and institutions’ efforts to make do with absences ‒ as accounting
to inscription work (Latour, 1987; Robson, 1992; Robson &
provides only limited understanding of complex organisational
Bottausci, 2018). The opening quotation illustrates the relation-
realities (Ahrens & Chapman, 2004, 2007; Chapman, 1997;
ship between accounting ‒ in the form of a forecast ‒ and the
Jørgensen & Messner, 2010) ‒ by stimulating dialogue (Busco &
multiplicity of a demand chain. The forecast, a central part of target
Quattrone, 2015, 2018; Quattrone, 2009, 2015) via mobilising in-
setting (Frow, Marginson, & Ogden, 2010; Hansen, 2012; Merchant,
stitutions (Dambrin & Robson, 2011), politics (Briers & Chua, 2001;
2006; Merchant & Manzoni, 1989) and budgeting (Ezzamel,
Chua, 1995), and other forms of resources (Ahrens & Chapman,
Robson, & Stapleton, 2012; Marginson & Ogden, 2005; Miller &
2007; Jørgensen & Messner, 2010); and by experimenting with
O’Leary, 2007; Preston, Cooper, & Coombs, 1992), here enacts the
different accounting devices (Andon et al., 2007; Chua & Mahama,
multiplicity of the demand chain ‒ a lateral ordering of the firm’s
2007; Skærbæk & Tryggestad, 2010). These studies focus on the
ways in which complementing work is made ‘highly compatible’
with incomplete accounting (Jørgensen & Messner, 2010, p. 188).
* Corresponding author. Our study brings a different focus to absence through a concern
E-mail address: [email protected] (L. Yu).

https://doi.org/10.1016/j.aos.2020.101129
0361-3682/© 2020 Elsevier Ltd. All rights reserved.

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
2 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx

with multiplicity. When an object multiple is enacted as a variety of processes of accounting in action’ (Hopwood, 1976, p. 3) where
durable but different practices, absences may take many forms ‘both a fluidity and a specificity have been introduced into our
(Law & Singleton, 2005; Mol, 2002), each of which may challenge understanding of accounting’ (Hopwood, 1987, p. 231).
the way it is represented by accounting. This sets it apart from the One sign of accounting-created struggle is the tension of
research on incompleteness and instability because contextualising incompleteness. Managers, in a pragmatic sense, may feel that ac-
accounting with complementary resources may not relate to the counting should provide a precise guide for decision making and
many different types of absences that come into tension with ac- action, but ‘accounting information usually does not capture all the
counting’s representation(s) of the object. In such a situation, dimensions of performance considered relevant for an organisation
relating the distribution of absences to the multiple enactments of or a manager’ (Jordan & Messner, 2012, p. 546). ‘Managers therefore
an object multiple via mediations of accounting is worth exploring. tend not to rely ‘blindly’ on such information. They rather seek to
This motivates the research question of this study: How does ac- contextualise or complement it by drawing upon other inscriptions
counting act on a demand chain that is practised as a multiple? or forms of knowledge’ (Jordan & Messner, 2012, p. 545). This
Our case study of EuroTech (pseudonym) examines how a complementing work around accounting has interested re-
forecast interacts with the challenges of lateral coordination be- searchers contemplating what is absent from accounting (Ahrens &
tween sales, factories, suppliers, and customers in managing de- Chapman, 2004, 2007; Busco, Quattrone, & Riccaboni, 2007; Jordan
mand chain. There are tensions in the demand chain between, for & Messner, 2012; Vaivio, 1999), and this literature examines, for
instance, sales and production, as shown in the opening quote, in example, how accounting is complemented by strategising
response to a market-based sales and operations planning (S&OP) (Jørgensen & Messner, 2010), by relating to strategic guidelines and
forecast. ‘Volumes sold’ is a sales activity while ‘shipment’ is a operational activities (Ahrens & Chapman, 2007), and by capital-
procurement and production activity; the practice of the demand ising on intimacy to customers (Vaivio, 1999). These interventions
chain in sales differs from that in production and the demand chain may help make the absences from accounting representation less
is enacted as an object multiple (Law & Singleton, 2005; Mol, 2002). worrying by a dialogical process that generates sensible and even
Thus, a general forecast may not have a stable relation to the de- creative ‘in-tensions’ (Busco & Quattrone, 2015, 2018; Quattrone,
mand chain and may not coordinate diverse entities across it; it 2017). Complementing work stimulates ‘doubts’ and ‘debates’
creates tensions. Its purpose to coordinate laterally interdependent that create more opportunities for decision making.
organisational functions (Atkinson, 2009; Tohamy, 2008; Vollman Another tension is instability. When accounting calculations are
& Cordon, 1998; Vollman, Cordon, & Heikkila, 2000) is at risk corrupted or flawed they require additional justifications. Dambrin
when the demand chain is practised differently in relation to the and Robson (2011) show how institutionalisation and bricolage of
market, customers, products, production, capacity, and suppliers. many inscriptions make it possible to accept the consequences of
This paper adds to the literature on inscription work by ana- even ‘flawed’ calculations so that, for example, bonus payment
lysing how managers find and organise frictions and objections systems may endure in spite of calculative flaws. Frandsen’s (2009)
(Latour, 1999b) generated from the multiplicity of an object. Here study of the origins of accounting relates the difficulties of tracing
accounting actively enacts e via discovering and coordinating the costs to the disease of psoriasis in a hospital. This problem of absent
differences in practising e an object multiple. The analysis is cen- (hospital) realities is also central to Chua’s (1995) study of what she
tred on the ways in which a forecast both provokes objections to its calls ‘flawed approximations’ of Diagnosis-Related Group (DRG)
inscription and transforms such objections into decision-making costs that seemingly had little other presence than political com-
activities. This is relevant because the challenge of acting upon an promises. The reality of ‘hospital business’ faded away while a re-
object multiple involves discovering and distributing objections via ality of ‘political business’ gained prominence. In another study,
mediations of accounting across its multiple enactments. Briers and Chua (2001) attend to the lack of decision coherence
To explore how forecasting acts on the multiplicity of the de- on costing, finding it was resolved, not by approaching the realities
mand chain, the paper is organised as follows. The next section of production of aluminium, but by reconciling them with mana-
discusses the literature on accounting’s precarious relation with gerial intuitions, alternative business, and system insufficiencies.
practices vis-a-vis incompleteness and instability, followed by a The problem of arriving at realities is about the instability of the
discussion of the implications of accounting’s engagement with inscription building process on the one side, and reluctance on the
material absences on the enactment of objects multiple. Then it other to make the realities of production a strong argument. A
develops the theoretical approach, based on modes of existence, similar problem with instability of inscription work is presented by
that informs the analysis. The paper goes on to describe the method Andon et al.’s (2007) analysis of the development of performance
used. The findings are presented in seven empirical episodes, fol- measures. Here, lack of coherence of performance management is
lowed by discussion of the study’s contributions and implications. attributed to a relational drift, which is accommodated through a
series of experimentations around accounting aimed at connecting
2. Accounting’s precarious relation with practices accounting to otherwise absent realities of ‘performance’.
Generally, accounting produces tensions because of things ab-
2.1. Incompleteness and instability sent in its representation. This absence provokes complementing
work, such as alternative practices, bricolage, politics, and experi-
Research into the relationship between accounting and mana- ments. That is, managers seek to ‘fill in the gaps’ by adding contexts
gerial practices has established that accounting rarely creates to accounting. Accounting helps to perform a space of absences
comfort. This happens because accounting is not a machine simply (Quattrone, 2017; Quattrone & Hopper, 2006). It proposes a course
supplying answers to questions, but rather a machine that brings to of action based on managers’ ability to be provoked or motivated by
the fore worries, tensions and conflicts (Burchell, Clubb, Hopwood, its message (Busco & Quattrone, 2015, 2018) and allows them ‘to be
Huges, & Nahapiet, 1980); ‘Accounting information e even if somewhat relaxed about the representational qualities of ac-
available in detailed form e provides only for a limited under- counting information’ (Jordan & Messner, 2012, p. 545).
standing and handling of the complexity of organisational life Yet, this requires that it is possible to retain a sense of ‘relaxa-
(Chapman, 1997)’ (Jordan & Messner, 2012, p. 545). In other words, tion’ (Jordan & Messner, 2012) toward absences via contextualisa-
it creates, rather than resolves, managerial struggles. Therefore, it is tion. However, when multiplicity is brought to the fore, relaxation
important to ‘study the complexities of the evolving dynamic is difficult because complementing and contextualising may not

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx 3

accommodate the interactions and objections between the many the impossibility of full representation. In their study of the con-
different absences of an object multiple (Law & Singleton, 2005; struction of Siena Cathedral from 1259 to 1357, Giovannoni and
Mol, 2002) and accounting’s rendering of them. The present study Quattrone (2018) relate the incompleteness of the cathedral to
considers an alternative, where accounting actively helps discover the impossibility of filling the void between contested readings of
what absences are material and responds to the frictions brought by its representations. This tension, which relates to inter-institutional
these absences. This alternative requires an approach that allows an disagreements, differs from Singleton and Law’s concern with
exploration of how accounting helps to enact an object’s multi- alcohol liver disease, where tensions are not negotiated. Tensions
plicity through engagement with material absences. emerge because the object is multiple rather than because a single
object is open to opposition as to what it is and should be; tensions
2.2. Materiality of absent realities exist because there are many (absent) practices e enactments e
within an object rather than because the object leaves a lacuna for
Some accounting research does focus on the role of material managers to fill debates. To understand how accounting acts upon
objects. For example, Chua and Mahama (2007) study how ‘buyer‒ multiplicity, it is therefore compelling to position accounting next
supplier relations’ responded to a new accounting technique, re- to the many material absences that make the object multiple.
opened discussion about relations with a supplier and changed One way to study accounting when confronted with multiplicity
accounting to suit a new set of relations. Likewise, Skærbæk and is to follow its participation in discovering and organising the
Tryggestad’s (2010) study of accounting and corporate strategy tensions and frictions e or to use Latour’s (1999b) term, objections
details a history of problems, in response to which a series of new e arising from multiplicity. As a situated practice (Ahrens &
accounting devices were constructed to deal with emerging con- Chapman, 2007), accounting may help enact an object’s multi-
cerns and issues requiring new accounting calculations. The fric- plicity. However, this requires the interactions between accounting
tions became strategic challenges that rendered existing strategic and the multiplicity of an object to be taken literally. In this regard,
propositions outdated, requiring new calculations for capital Latour’s (2007) distinction between modes of subsistence and
expenditure. These contributions offer accounts of how accounting reference is helpful: Here both accounting and objects move and
adapts to other objects but pay less attention to the ambiguity of sometimes these movements intersect. These intersections pro-
objects, to their fluidity (de Laet & Mol, 2000) and multiplicity (Law duce objections and reformulations of what accounting accounts
& Singleton, 2005; Mol, 2002). Here, the various enactments of an for.
object may challenge each other, questioning accounting’s capacity
to establish faith in the way that decision mechanisms can hold 3. Subsistence and reference as modes of existence
together different absentepresences (ibid.) within an object
multiple. Studying relationships between accounting and the demand
Mol (2002) and Law and Singleton (2005) elaborate the notion chain as a multiple requires observation of episodes where the
of an object multiple in their studies of atherosclerosis and alcohol multiplicity of the demand chain is enacted through
liver disease respectively. Each of these diseases is enacted as absentepresences and what accounting reveals about them. To
multiple practices. For instance, alcohol liver disease is a lethal help gain access to material absences, Bruno Latour’s distinction
condition that calls for abstinence in the hospital; it is a problem between modes of subsistence and of reference is a relevant
calling for regulation in the substance abuse centre; it is something methodological tool. He claims that realities are enacted when
that is at least better than recreational drugs in GP’s surgery; and it knowledge of objects intersects with subsistence of objects, both of
is an effect of accumulation of other social problems in community- which have histories (Latour, 2007).
based social care (Law & Singleton, 2005). Therefore, ‘enactments’ Latour’s (2007) study of the evolution of horses featured in the
of an object multiple ‘take place in the practices of getting to know Natural History Museum in New York offers an illustration. He re-
those realities’ (Law & Singleton, 2005, p. 334), as each ‘produces visits the confusion attributed to the distinction between the his-
their object in question’ (ibid., p. 336; see also Mol, 2002). Central to tory of objects and the history of knowledge about objects (history
this ontology multiple is a set of absent‒presences (Knox, of science). The classic theory states that the history of horse
O’Doherty, Vurdubakis, & Westrup, 2015; Petani & Mengis, 2016), lineage forms ‘a simple evolutionary sequence: from small to large
for example, the presence of the treatment of the disease in the bodies, from many to fewer toes and from short to tall teeth’
hospital requires absence of alcohol. Therefore, the presence of the (Latour, 2007, p. 4). A more recent theory, however, indicates that
disease is ‘generated in juxtaposition with realities that are horses’ evolutionary path is more complicated, based on the recent
necessarily absent, even though they bring versions of those re- discovery and study by palaeontologists of fossils: Some horses are
alities to presence’ (Law & Singleton, 2005, p. 345). smaller than their ancestors and some still have three toes instead
Multiplicity becomes a problem when some enactments chal- of one. Evolution apparently takes a bushy rather than linear path.
lenge others. This is highlighted by a community-based psychiatrist The museum’s curators used the phrase ‘we now know’ to indicate
discussing alcohol liver disease: ‘it is not just a question of being the advancement of knowledge. The puzzle of how scientists
substance-free. It has to do with improving other aspects of life … gained more knowledge about horse lineage made Latour re-
enjoy health and a social life’ (Law & Singleton, 2005, p. 345). contemplate the distinction between objects and knowledge of
Practising the disease in the hospital is challenged by practising the those objects. To Latour, the curator’s careful presentation of a more
disease in social care. There are interactions and objections be- complicated version of knowledge of horse evolution and their
tween material absences: alcohol, as a material absence in the genuine statement of ‘we don’t know for sure’ (Latour, 2007, p. 4)
hospital, is in tension with the absence of ‘other aspects of life’ in does not confuse visitors and force scientists to ‘abandon all hopes
social care. Therefore, multiple enactments raise a challenge of to know something objectively’; ‘Quite the opposite’ (p. 6).
coordination. Enacted realities are not independent; these are not Knowledge about horse evolution becomes more objective. In
parallel worlds but different e sometimes competing e realities of contrast to the classic representational scheme where knowledge
an object multiple. and objects sit at opposite poles, Latour (2007) opts for an alter-
The tension related to material absences has been addressed by native methodology. In his ‘continuous scheme’, there is no
Giovannoni and Quattrone (2018), who understand the problem of knowing subject, ‘representation’, or ‘idea’. Nor is there a bridge
coordinating absences as an inevitable consequence derived from between the object and the knowing subject. Instead there is a

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
4 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx

history of subsistence of objects (e.g., fossils) and a history of significant revenue that would have been generated had factories
reference of the subsistence (e.g., scientists’ statements about horse had sufficient capacity. Inventory was a notable absence. In order to
lineage). Knowledge is acquired when these two histories intersect, balance demand and supply, GDC decided to implement a Sales and
which they do multiple times. According to Latour (2007), ‘the Operations Planning (S&OP) process with ID as a pilot site and
main problem of knowledge is to deploy the continuous chain of aiming to transform the company into a demand driven organisa-
experience to multiply the crossing points at which it will be tion. The S&OP process proposed that sales calculate a market-
possible to retroactively decide whether we had been right or driven S&OP sales forecast to be used by factories for their prod-
wrong about a given state of affairs’ (p. 16). uct line planning activities. However, many objections from the
In Latour’s proposal, horses existed in a reproductive form many demand chain made this proposal difficult and challenged the
years ago in the wild. It is impossible to observe how these ancient proposed seamless lateral coordination in a variety of ways. Though
horses lived and evolved, but they leave fossils as instantiations, as challenged, the forecast was trusted by the GDC for its capacity to
items of subsistence. When these fossils are ‘unearthed, trans- enable lateral coordination of the demand chain. Of interest to the
ported into crates, cleaned up, labelled, classified, reconstructed, research is how the forecast added to the multiple enactments of
mounted, published in journals’, that is, ‘once palaeontologists have the demand chain.
crossed path with the ancient horses’ (Latour, 2007, p. 24), a
knowledge of horse lineage is produced (p. 8). At each crossing 4.2. Data collection
point, modes of subsistence and reference intersect and interact,
and a facet of an object is enacted. Latour’s (2007) point is that Most of the empirical data was collected from documents, in-
neither the mode of subsistence nor the mode of reference is terviews, and observations of meetings pertaining to or held by
immutable over time. The challenge is to find more crossing points. actors relevant to the S&OP sales forecast. One of the researchers
The notion of crossing points is relevant for this study as in which entered the field in June 2010 when GDC was rolling out the pilot
we consider the relation between an object such as the demand S&OP process in some of the product groups in ID. Upon
chain (or ancient horses), a series of items of subsistence such as completing the field research in December 2011, the pilot S&OP
product, customers, sales, and market (or fossils) and under- process was still under construction. This researcher was present in
standing these through forecasts (or the non-linear theory of the the company interviewing and observing during the period from
horse lineage). Like the galloping wild horses, the demand chain is June 2010 to December 2010. After this point in time he did a
inaccessible in its entirety due to the impossibility of full repre- number of follow up interviews when new concerns emerged, up to
sentation (Giovannoni & Quattrone, 2018). It exists for managers in December 2011.
episodes, as will be narrated later, manifested by tensions around The empirical field involved the company’s headquarters and
supply shortage, composition of the market, product and customer adjacent factories. The researcher conducted 41 semi-structured
variations, planning rhythms and so on, all of which add materi- interviews (including six telephone interviews due to inability to
alities and realities to the demand chain, thus enacting its multi- conduct face to face interviews) with 16 managers across functional
plicity. Crossings between subsistence and accounting are access groups during the period June to December 2010 and thereafter
points to episodes in the enactment of the demand chain, which intermittently to December 2011. Each interview lasted between
‘take place in the practices of getting to know those realities’ (Law & 0.5 and 3 hours with an average length of around 70 minutes and all
Singleton, 2005, p. 334); accounting in the form of forecasting were tape recorded and transcribed. The researcher either partici-
helped managers ‘get to know’ the ‘realities’ of the demand chain pated in, or conducted interviews with participants of, the pilot
(Law & Singleton, 2005). This methodology does not describe all S&OP product group planning (PGP)1 meetings. Each meeting took
possible realities but only those enacted at crossing points; how- around 2 hours and was tape recorded and transcribed. A vast range
ever, it does help researchers identify intersections, which become of internal materials was studied including S&OP charters, 6 Sigma
empirical observation points to access the demand chain multiple. charters, business cycle forecasts, financial forecasts, an S&OP in-
struction manual, data in the pipeline and in the Demand Solutions,
4. Methodology: studying accounting and enactments of the factory daily planning inscriptions, factory stock levels, safety stock
demand chain multiple levels, shipment histories, and meeting minutes. As the S&OP
process was implemented in ID, most of our data is collected from
4.1. The S&OP process at EuroTech interviewees, meetings, and documents within ID and GDC. Table 1
shows the summary of the empirical data collected.
The field study was conducted at a large European high-tech
manufacturing company, EuroTech. Its product range comprised 4.3. Data analysis
five platforms including bearings, seals, lubrication systems,
mechatronics, and services. Its customers were from a wide range The organisation and analysis of data focused on crossing points
of industries including agriculture, automotive, construction, elec- e episodes where the demand chain via intersections between
tric power tools, home appliance, oil and gas, industrial pumps, subsistence and references gained more properties. Empirical evi-
green energy, and so on. There were three divisions: the automo- dence was organised by crossing points between items of subsis-
tive division (AD), industrial division (ID), and service division (SD), tence (mode of subsistence) and forecasts (mode of reference),
each of which served distinctive groups of customers, and business which helped to articulate the demand chain as a multiple. Each
volumes varied. Each division’s products were part of the five crossing point concerned a particular problem such as ‘supply
platforms mentioned above but the products manufactured and shortage’, ‘composition of the market’, ‘product and customer
sold varied. Each division had a manufacturing organisation and a variations’, and so on; each denoted a tension due to the intersec-
sales organisation. In coordinating these three divisions, Group tion between reference and subsistence.
Purchasing supervised material sourcing and Group Demand Chain This procedure focuses on the emergence of the demand chain
oversaw planning across the demand chain. Fig. 1 illustrates
EuroTech’s structure.
In 2007, Group Demand Chain (GDC) realised that EuroTech had 1
A Product Group Planning (PGP) meeting raises concerns about planning issues
declined a substantial number of customer orders, missing out on for a number of different product lines.

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx 5

Fig. 1. Organisational structure of EuroTech.

Table 1
Summary of interviews, meetings and documents relevant for the study.

Interviews

Positions Face to face Telephone

Sales Manager of Bearings and Units e ID 1 0

Demand Chain Manager for large bearings e ID 5 2
Manager on Manufacturing & Supply e ID 5 1
S&OP Manager e GDC 4 0
Business Process Analyst A e ID 1 0
Business Process Analyst B e ID 3 0
Business Process Analyst C e ID 2 1
Regional Sales Director e ID 1 0
Sales Manager e ID 2 0
Product Line Planning Manager A e ID 1 0
Product Line Planning Manager B e ID 2 2
S&OP Product Line Planning (PLP) Manager e GDC 3 0
S&OP Supplier Capacity Planning (SCP) Manager e GDC 2 0
Forecasting Manager e SD 1 0
Purchasing Manager e ID 1 0
Director of Demand Chain e GDC 1 0
Total 35 6

Meetings Attendance

Pilot S&OP meeting Sept. 2010 1

Pilot S&OP meeting Feb. 2011 1
Total 2

Internal documents

S&OP charter
6 Sigma charter
Business cycle forecasts (the F18 curve)
ABC analysis
Financial forecasts
S&OP instruction manual
Pipeline
Demand solution
Factory daily plans
Factory plan, stock levels & safety target levels
Shipment histories
Meeting minutes

as an object multiple. It does not guarantee that all practices have 5. Intersections between modes of subsistence and reference
been discovered and that the research has found all crossing points. on the demand chain
Yet, this is not a constraint. The theoretical reason for this is that the
study emphasises how accounting actively enacts and responds to The following analysis of crossing points shows first how the
multiplicity. The methodological reason is that following crossing development of the S&OP forecasting mechanism was both
points that organise the forecasting work in the firm helps to an- conditioned by and enacted the discovery of items of subsistence of
chor data collection. This makes it possible to consider who should the demand chain. It also shows how it translated objections from
be respondents and where to go and to observe (next). Therefore, subsistence into accounting in the form of decision mechanisms for
there is an interaction between the material absences of the de- coordinating the multiplicity of the demand chain.
mand chain and follow ups of data collection. This means that our
theorisation is always in conversation with the literature, analysis 5.1. Crossing point 1: supply shortage
of collected data, and collection of new data.
When asked why the S&OP process was proposed, many

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
6 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx

managers pointed to availability failure, for instance: requirements. Each factory has been going to their own sup-
pliers, (but) our purchasing organisation didn’t have a detailed
We have poor figures on more or less all channels … this is the
organised updating of the information of the demand of
actual situation when it comes to deliveries right now, so that is
different components, materials and so on … there wasn’t
not a good picture now … I just give you a hint on our avail-
anything structured already in place.
ability. Here (pointing at a computer screen) you see the H
channels that we have, you see who the planner is, here we have
the availability on the stock items, 81% for the H2, 9, 35, 51, 53 The presence of many calculating agencies exacerbated delivery
and so on. (Purchasing Manager e ID) problems. Sales predicted market demand, factories anticipated
production runs and batches, and the purchasing organisation
We have a factory in India now, which has availability failure of
struggled to make contracts with suppliers; each developed their
50%. I mean it’s just ridiculous. I mean it doesn’t matter because
own model of planning and communication and the three func-
the sales guys don’t talk to customers anymore. (Business Pro-
tions faced different product and customer priorities. The multi-
cess Analyst B e ID)
plicity of calculating agencies increased the complexity and the
unpredictability of the demand chain. These various items of sub-
When subsistence such as ‘empty’ inventories and ‘lost’ cus- sistence identified these complex availability problems as due to a
tomers emerged, the demand chain was understood as availability. lack of general coordination between operations (supply) and sales
These items of subsistence proposed the firm as ‘not being able to (demand). To manage this coordination concern, a reference was
predict increasing demand early enough’ (S&OP Manager). This produced to balance demand and supply: ‘we want to compare our
imbalance between demand and supply was also connected with capacity to our (S&OP) sales forecast three years from now so that
another subsistence in product delivery, as explained by the S&OP we can increase our capacity’ (Product Line Planning Manager A ‒
Manager: ID).
To re-capture lost customers the S&OP sales forecast was pro-
In EuroTech we produce, for example, a bearing which is mainly
posed as a market-driven unconstrained forecast. Business Process
for car customers (customers to the automotive division), but
Analyst A e ID said that this ‘pure customer demand’ would enable
there is always a certain part which goes to either service divi-
‘decisions and actions before everything is a big mess’. Or as pro-
sion or industrial division. So, it’s very rare that a product only
posed by Product Line Planning Manager B e ID:
goes to one segment. Factories produce both for particular
customers (for a particular division) and for (all) EuroTech I’m really positive towards this S&OP forecast because we
customer segments. That’s why it’s so important to get the (product line planning in factories) will have to communicate
whole demand right for all our customers … We had different with sales. I’m sure it will improve cooperation between sales
solutions, pieces here, pieces there. and manufacturing, but it will probably take some time … We
actually had one case when we had one final variant where the
sales side says we are going to sell 2000 per month and we (the
This quotation attributes the issue of complex delivery to the
factory) had only shipped 1000 per month, and we only had
lack of coordination along the demand chain. Each component
orders for 1000 per month. Then we ask them (sales), OK, we
could be delivered to customers who ordered from all three di-
had only sold 1000 pieces per month, you say we are going to
visions in EuroTech. This delivery complexity escalated because of
sell 2000 pieces per month, either you have to place more or-
the platform concept as explained by the S&OP PLP Manager e
ders, or to decrease the forecast.
GDC,
There are five different platforms (product groups). We have
Aiming to address the tension between demand (customers)
bearing units, which is the largest one. There are also seals,
and supply (inventory and delivery) EuroTech proposed an un-
lubrication systems, mechatronics and services … The main
constrained market-based S&OP sales forecast for the upcoming 36
purpose is to combine as many of these platforms as possible.
months as a solution. Its purpose was to integrate sales, production,
The more we can combine these solutions, or platforms, the
and suppliers to reach a consensus production plan so that Euro-
better it is for us and for the customers; and it also gives more
Tech would increase investment in capacity if predicted market
value to us and customers.
demand exceeded current capacity. A consensus plan would
potentially bring every party’s information, action, and communi-
As the quotation explains, EuroTech delivered not only one cation flow into one frame where product delivery could be well
product to multiple customer segments but also combinations of organised, empty inventories reduced, lost customers re-captured,
product groups (platforms) to one individual customer. For and availability increased. A customer friendly supply would create
instance, it could sell a combination of bearings and mechatronics proximity (Corvellec, Ek, Zapata, & Zapata Campos, 2018) between
to one customer as something managers called ‘a solution’. This all these items of subsistence.
delivery complexity was an item of subsistence of the demand
chain that put a burden on lateral information, communication, and
5.2. Crossing point 2: composition of the market
action flow. The early quotation from the S&OP manager, ‘we had
different solutions, pieces here, pieces there’, was further explained
However, with this market-oriented solution, an objection from
by the S&OP SCP Manager ‒ GDC,
the demand chain emerged. Forecasting of the market had to be
All this information (forecasts) has been flowing to the factories, made at the ‘different levels’ (Forecasting Manager e SD) because,
but never in a unique, organised structure. Factories are of for example, ‘business with a huge automotive client (of AD) and (a
course planning their production and capacity again with a client purchasing) smaller bearings (from ID and SD) are certainly
different methodology (to that of sales). And on the supplier different’. Therefore, the many types of products offered and the
side, of course all factories are communicating their supply divisional structure of EuroTech made forecasting of the market
needs. Yet, there has never been a consolidation of these difficult.
The unconstrained market-based forecast (the reference)

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx 7

created at crossing point 1 was not able to relate the heterogeneity proposition of an S&OP sales forecast in the same meeting:
of customers to the market. This absence emerged in the inter-
Normally, we (EuroTech) have a better understanding of what
section between the forecast proposed at crossing point 1 and two
trends and thus the future will be than some of the customers
competing sets of subsistence of the demand chain at crossing
that we asked, so what is needed is to apply a certain top-down
point 2. One set of subsistence, proposed by GDC, suggested the
logic to say, OK, where is the general trend, can we anticipate
market as made of customers. Thus, the market would be aggre-
things because customers don’t know yet … you cannot only
gated from either the lower customer item level, such as a partic-
work on the detailed bottom-up forecast because the truth is
ular bearing sold to a particular customer, or the higher product
not there to be caught …
line level, such as a particular size of bearings that could be used by
many customers. A customer item and a product line would be two
possible ‘primary keys’ for the forecast, the importance of which is The quotation indicates two competing sets of subsistence:
illustrated in the following quotation by the S&OP Manager and Customers’ own voices versus economic conditions. The competi-
summarised by Table 2: tion was conditioned by the difference between what the cus-
tomers said and did. When intersecting with the unconstrained
Automotive Division (AD) had only a few big customers, large
market-driven forecast developed at crossing point 1, each
organisations, each of whom ordered a large number of prod-
claimed a different type of knowledge for governing the demand
ucts. They prided themselves of sound supply chain manage-
chain: The bottom-up method took advantage of relations with
ment (SCM) and expressed certainty about future business
customers, and the top-down approach took advantage of macro-
volume. Here, the primary key could be set at the lower
economic knowledge developed by a central business planning unit
customer item level. In Industrial (ID) and Service Divisions
in EuroTech. The question became: Who would be a better
(SD), in contrast, the order book for customers had a short time
spokesperson for the market ‒ individual customers or general
horizon, if any. They had a large number of smaller customers
market conditions?
and customer relationship management consisted of personal
The items of subsistence of the demand chain at crossing point 2
relationships between salespersons and customers, and
comprised customers’ voices, deviation between what customers
customer demand was understood as unpredictable. In this
said and did regarding future business volume, and general eco-
situation, the primary key was set at the higher product line
nomic trends. They were absent at crossing point 1, but they were
level.
materialised because of an objection to the market-driven forecast
generated at crossing point 1. The intersection of subsistence and
In spite of their differences, AD and ID/SD proposed a common reference presented a new tension to the managers: Specific
strategy to forecasting based on a primary key as anticipation of knowledge of customers would not be compatible with general
individual customers’ business volume, which would then aggre- knowledge of market trend. This raised a new concern about co-
gate up to regional sales volume. ordination: How were customers linked to the market? What was
Such a bottom-up forecast relied on existing customers’ past the composition of the market?
behaviour to anticipate their future demand. This approach, how- There was thus a movement from inventory planning, logistics,
ever, revealed another subsistence of the demand chain: The dif- and calculative practices to customer relations, customers’ SCM,
ference between what customers said and what they did, as order books, divisions inside EuroTech, and general economic
suggested by the Sales Manager of Bearings and Units e ID during a outlook; a movement from enacting the demand chain as a tension
pilot S&OP product group planning (PGP) meeting, between demand and supply to that of enacting it as a composition
of the ‘market’. The unconstrained forecast was only a comforting
The thing about the detailed bottom-up forecast is … our
reference at crossing point 1, and it spun into two possible modes of
strength in sales so that we can ask the customers about their
reference, which generated competing interventions at crossing
plans, and we can get more details. But of course each customer
point 2.
is a little bit inaccurate … They (salespeople) can only ask the
The bottom-up approach was eventually preferred not only on
customers who tell what they think.
its potential merits but also because of historical ignorance of the
availability issues of its alternative, the business cycle model, even
The tension was built on the contested inscriptions of the though sales had more local knowledge of what customers said and
market between GDC and sales. Sales questioned GDC’s assumption did than GDC. This, however, did not mean that the BCF exited the
that past experience about customers would be able to predict their demand chain. The bottom-up S&OP forecast did not replace the
future dealings with EuroTech; for sales, a business cycle forecast top-down BCF, rather the latter continued operating for its central
(BCF), which started from macroeconomic trends, would better planning purpose. Also, the BCF did gain some prominence in
account for the market. This would be a top-down forecast based on product line planning in factories, as cautiously mentioned by the
macroeconomic indicators, not on what individual customers Demand Chain Manager of Large Bearings e ID,
might say. The manager explained the merits of a BCF against GDC’s

Table 2
Properties of actors in divisions that created two different primary keys.

Actors Automotive Division (AD) Industrial (ID) and Service Divisions (SD)

Business characteristics Large volume of products sold to few number of customers Low volume of products sold to a large number of customers
Order Book Long and reliable Short and unreliable
Degree of certainty of future customer volume High Low
Primary Key (mode of reference) (lowest) Customer item level (higher) Product line level

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
8 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx

We plan by looking at the last 12 months and the biggest weight value, and customer heterogeneity. For some products, the varia-
is the last 3 months … then we see, based on the business tion in the historical performance of products was substantial. On
development (the BCF), what they say where the business is managing this kind of variation, the S&OP manager pointed out,
going, then we put on some percentage because, for example,
We give all those cases a certain tag, so for example, if we only
now we are coming to an upturn of the business.
have order book, but no history, which means this is a new item,
we give a pre-warning as we never sold that item to any cus-
This started developing accounting as simultaneity (Latour, tomers. Now we (also) have the order book, so for sales persons,
2005) instead of as progression or succession: enactment of the it’s good for them to get these pre-warnings.
demand chain at this crossing point did not substitute the enact-
ment at crossing point 1. In addition to BCF, the reference of a 36-
Product novelty was an item of subsistence that required sales
month unconstrained forecast was still in operation to ensure
people to manually forecast items that had no history in the com-
comparability of all forecasts in sales and operations planning. At
puter but only appeared in the order book.
the series of pilot S&OP PGP meetings, all forecasts were adapted to
There were yet other items of subsistence as explained by the
a time horizon of 36 months. The BCF, used to inform product line
S&OP manager as he flipped to another Demand Solutions screen,
planning as indicated in the above quote, was already an uncon-
strained forecast estimating market trends. Therefore, the reference Or you have a strongly growing item, or your order book is much
created at crossing point 1, an unconstrained forecast for the up- bigger than your forecast … Here we have a decision tree which
coming 36 months, operated simultaneously in GDC, sales, fac- says in this case the sales per month is above certain amount of
tories, and the central planning unit; the additional model of money, so it is an important item, and we filter out items that
relating different customers to the market became useful for the are strongly growing, of course it’s a question of how you define
calculation of an S&OP forecast for coordinating the demand chain. a growing item. We have two definitions, one is year over year,
All co-existed and were added to the demand chain making it more so last 12 months should be 50% above the year before, and the
multiple. last quarter needs to be 100% over the quarter the year before …
Here is the situation where we have last 12 months, before
certain period of time, it was nothing. We have according to that
5.3. Crossing point 3: product and customer variations
definition a growing record. … And combine that with those
filtered events, then you have a very powerful tool …
At crossing point 2, individual customers were taken seriously;
the bottom-up reference was established. However, this also
created an additional objection from the demand chain, which the Product growth was another item of subsistence. Again, human
Manager of Manufacturing & Supply e ID referred to as the amount actors would have to forecast items with growth that exceeded a
of forecasting work created because of the ‘sheer number of certain threshold manually. To this, a Forecasting Manager e SD
different products’ sold to ‘very different groups of customers in later added that ‘we are also presenting average sales value and
each division’ (AD, ID and SD). those items with high value will be forecasted manually’. This
Forecasting actors e salespersons, sales managers and sales revealed a fifth item of subsistence in the form of product value.
directors e were unable to calculate forecasts for the whole variety Customers’ voices were heterogeneous because when the
of customer-items and product lines manually because of limited bottom-up customer-based forecast intersected with items of
time. Instead, part of this work was delegated to a software in- subsistence such as product novelty, product growth, product
strument called Demand Solutions. This software compared histor- value, and so on, different types of forecasting models were pro-
ical forecasts and actual sales for the last 12 months and predicted posed. Each prioritised certain customers and products; each
sales for the upcoming 36 months. The S&OP Manager spent an related customers to products in a unique way. In an attempt at
afternoon demonstrating Demand Solutions to the researcher, making heterogeneity a little less prevalent, a reference was created
explaining how it created different inscriptions that linked cus- that ranked products on revenue into A, B, C, D, and E items (those
tomers to products through different divisions of labour. contributing 30%, 30%, 20%, 15%, and 5% to total revenue respec-
The use of Demand Solutions, however, was not without obsta- tively). Human forecasting agents could influence A, B, and C
cles, as highlighted by the S&OP Manager, products by forecasting them manually. While A, B, and C products
contributed to 80% of turnover, they were related to only 3% of
If we take out total automotive business worldwide (AD), we are
customers. Therefore, the reference constituted a hierarchy in
talking about 5,000 records in combination with final customers
which products and customers were prioritised in the demand
… If we take the same (primary) key for the SD in Europe, we
chain.
have 3 millions of these records e items and final customers (He
This crossing point enacted yet another materiality of the de-
was scrolling down the screen to show the massive size of the
mand chain in relation to the tension of exacerbating forecasting
records. Then he clicked on one item.) On that (product) line,
complexity due to (too many) types of customers and (too many)
you will have something like this, very erratic sales patterns, 5
variations in products in the three divisions of EuroTech. Items of
pieces here, 20 pieces there. Here for a lot of months, nothing
subsistence such as product volumes, growth, novelty, value, and
(has been sold), so what the system (computerised) forecast
customer heterogeneity emerged and objected to the bottom-up
creates will be very bad.
forecasting approach constructed at crossing point 2. This tension
presented a coordination concern for managers: How do we relate
The number of customer records, the differences between cus- different types of customers to the ‘sheer’ variations of products?
tomers belonging to different divisions, and erratic sales patterns The different visualisations in Demand Solutions, from order book to
made it difficult for Demand Solutions to generate a sensible the ABCDE system, helped bundle customers and products differ-
bottom-up forecast developed at crossing point 2. When meeting, ently. For instance, sales people planned for large customers
and objecting to, such a customer-oriented (bottom-up) forecast, ordering a few big items, as well as new products and products with
the demand chain revealed items of subsistence, as shown below, high value and growth. Other customers and products were
such as product volume, product growth, product novelty, product

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx 9

managed through the software. These practices took place generate the smallest variance compared with actuals. This was
simultaneously. termed ‘error’ by managers. Yet, error existed in many forms, two of
which were relevant here. The so-called absolute error e the ab-
solute difference between actual and forecasted sales e was
5.4. Crossing point 4: average customers calculated on raw data. If actual sales were 100 and forecasted sales
were 70, the absolute error would be 30. This calculation was
As a consequence of handling product and customer variations, automatically built into Demand Solutions and therefore the system
the forecast became very detailed at crossing point 3, and this would, by default, choose the formula that would generate monthly
raised an objection from GDC, namely that (too) many details did sales forecasts for the next 36 months with the smallest absolute
not produce a set of smooth monthly numbers preferable for me- error for the previous 12 months. However, as indicated in the
dium to long term capacity planning in factories, as exemplified by above quotation, absolute error did not produce ‘smooth’ monthly
the Business Process Analyst B e ID, forecasts. Instead, ‘mean error’ did. Mean error was the average of a
set of absolute errors. So, if there were absolute errors of 30, 10, and
You don’t want the (forecasting) system to be that easy picking
35, the mean error would be 25. This measure would produce a set
up the specific situations. We look for investment in the next
of smoothed forecasts, which were suggested as more in line with
three years. You can smooth the forecast, (say) your forecast is
S&OP’s time horizon for capacity management. Thus, erratic cus-
not going up by 10% but only by 5%.
tomers were translated into average customers. This required
forecasting managers to override Demand Solutions, which calcu-
The model created at crossing point 3 raised a dilemma about lated forecasts using the absolute error. To do so, managers
the link between investment in production and calculation of ‘exported (historical) data from Demand Solutions to their own excel
customer needs. If a forecast was too ‘precise’, each monthly fore- spreadsheets’ where ‘mean error’ could be generated.
cast would be unique. This made customers erratic, which became a This further multiplied the number of simultaneous forecasts.
barrier to medium to long term capacity planning in production, as The forecast created at crossing point 3 was not replaced. It still
proposed by GDC. Capacity planning did not ask questions about functioned as a mechanism to relate customers to products. It
specific customers but about general operations: ‘do we need to existed in Demand Solutions to which sales, GDC, and production all
invest in the (production) channel?’, ‘do we need to invest in the had access. The ‘smoothing’ work was addressed by creating an
machine?’ or ‘do we need to build up a shift?’, as explained by the additional space outside the software. This simultaneity resolved
Business Process Analyst B e ID. Such questions were not attached the tension between precision in predicting customer wants and
to the S&OP forecast. Therefore, fluctuating monthly bottom-up generality in capacity planning.
forecasts, which created erratic customers, were in tension with
the subsistence of medium to long term production and machinery
investments. This presented a new concern for managers: How do
we relate production to sales? To address this concern, a model 5.5. Crossing point 5: planning rhythms
relying on average customers was proposed as follows by the
Business Process Analyst C e ID, Factories, on the other hand, had problems with the ‘smooth’
forecast because their planning horizon was short term product
… if you have a strange history, we have for example for large
line planning where ‘precision’ was a calculative must. This became
bearings, we had a situation a couple of years ago where you can
apparent when Product Line Planning Manager B e ID complained
wait for 1.5 years for your orders, which means you get no
that ‘availability for every material family now’ was more impor-
supply, no sales, no sales, no sales, and then suddenly there is a
tant than that ‘in three years’ time’. There was a tension between
production, and you produce all sales in one order, 20 large
operational product line planning in factories and medium to long
bearings in July, then there is a long period of no sales, no sales.
term capacity planning proposed by GDC and quantified through
Then of course we use that input to plan the forecast, but it is not
smooth forecasts. Average customers were at odds with product
really good. So in those kinds of cases, it is very valuable to
line planning. This tension first emerged with an item of subsis-
adjust the history to smooth out what the real one (history) is.
tence called ‘material families’ (product batches) as explained by
the S&OP PLP Manager B e GDC in relation to factories’ planning
This means that ‘adjusted’, ‘smooth’ forecasts were preferred to priorities as follows:
‘real’ ones. An average customer would be preferable to an actual
If we produce the material family first we have to look at the
customer for medium to long term capacity management. The
order book, so customers can get what they order. So, this is the
S&OP Manager explained:
first priority, and the next one is the practical distance from the
I believe when we talk about forecasting accuracy, it may be safety stock. If the distance from the safety stock is below certain
more important to select the formula with lowest mean error level for a particular material family, we need produce this
compared to the lowest absolute error because that one (mean material family first. So we look at which material family has the
error) will select the formula which produces smooth (monthly) worst situation.
forecasts.
This shows that, for factories, the first step was to use the order
Rather than ‘adjusting’ and ‘smoothing’ the detailed represen- book to determine which material family to produce first so that
tations of the customer-products relations modelled at crossing factories would have inventories available to fulfil these customer
point 3, GDC opted for smoothing the fluctuating monthly forecasts orders. The next step was to compare the stock level with the safety
by looking for calculative space other than Demand Solutions. The stock level for each material family. If the gap between the two for a
S&OP Manager referred to formulae embedded in Demand Solutions particular material family exceeded a certain level, then that ma-
to assist sales finding average customers. Demand Solutions could terial family was to be produced first. When explaining the
generate 21 different forecasts; in principle, the forecast to be importance of a material family, the Product Line Planning Manager
selected would have to be the one, it was argued, that would A e ID provided the following illustration,

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
10 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx

We have material families. Typically, it’s the size of the bearings. these items of subsistence of the demand chain and the reference
If it’s one size, it means you use one type of outer ring and one developed at crossing point 4, the ‘smooth’ forecast used to enact
type of inner ring. Then you can have different variations with the demand chain as medium to long term capacity planning,
different balls and different cages as well as other variations. revealed the tension on planning rhythms between factories and
There is also a decision (called an M decision) linked to that. GDC; average customers were at odds with daily product line
Here you have the main variant. Then when you have all the planning. It raised a concern about coordination of the demand
material, this is the type we want to produce and when this chain to managers: How do we organise the planning rhythm in
comes to the factory, you can turn it in different ways, you can production? Daily availability was top of the factories’ product line
have balls in the inner ring … you make some variations in the planning agenda whilst the average customer was related to their
turning. This is what we call the turning variant (called a D medium to long term capacity planning. This average customer
decision). Then we have final variants. You can have difference reference met objections from items of subsistence such as product
in clearance, the clearance between the rings (called an E de- batches and work-in-progress inventories, which were enacted
cision). M is (the decision) for (producing) the main variant, or ‘daily’, even ‘every moment’ as highlighted by the quotation. Here,
the material family. D is (the decision for producing) the turning the temporal and spatial organisation of postponement, where a
variant. E is (the decision for producing) the final variant. product was put together as late as possible to fulfil a customer
order quickly, organised production sequences. Daily product line
planning strived for optimisation within capacity constraints. This
A material family was a batch of bearings of the same size of
optimisation was achieved through references in factories such as
outer and inner rings. Factories made M decisions about which
order book, safety stock, and factory forecasts. However, these did
material family to produce first. In each material family they
not replace the ‘smooth’ sales forecast generated at crossing point
decided which turning variant to produce (D decisions). Finally, E
4. The smooth sales forecast retained its status in relation to me-
decisions concerned the final variant factories were to produce.
dium to long term availability. The decision model for medium to
This is why the M, D, and E decisions were called a set of product
long term capacity planning of the demand chain in GDC and sales
hierarchy decisions in EuroTech. D and E decisions were postponed
co-existed with the model for daily product line planning of the
as much as possible because the demand chain would be more just-
demand chain in factories.
in-time when the time gap between factories making E decisions
and customers placing an order was smaller. These product hier-
5.6. Crossing point 6: capacity constraints
archy decisions were introduced to mediate the relation between
availability and flexibility, as explained by the S&OP PLP Manager B
Crossing point 5 emphasised daily optimisation of product line
e GDC,
planning within capacity constraints and challenged the S&OP
So what they do is that they keep the full quantity open as long assumption that a forecast should be unconstrained. As the Man-
as possible for all possible variants. Then the order comes in and ager of Manufacturing & Supply e ID exclaimed, ‘factories did not
then we have total availability. Let’s say if we have 5 days’ lead consider (using) the (S&OP sales) forecast because they always had
time on the material (an M decision), and then you can take this constraints’. Production wanted to dominate sales because avail-
(D) decision in 15 days, and then when you are closer to the ability to existing customers was always important, but the market
actual production date you know more about whether it’s gonna kept growing and new customers joined. Excess and potential
be a tapered ball a cylindrical ball because things may change. customers emerged as items of subsistence that objected to the
You have increased the flexibility. factory planning models. When the S&OP process related sales to
operations planning, customers would be in excess for production,
and decisions would have to be made to exclude a certain demand
Here the mode of subsistence of the demand chain in factories
from the S&OP process. Therefore, a tension emerged in which sales
emerged as the product hierarchy, work-in-progress inventories (D
wanted to include excess and potential customers in the demand
and E decisions in each material family), and product batches (M
chain whilst factories would only relate production to existing
decisions), all aiming at flexibility and availability. At stake was the
customers. This tension created a new coordination concern for
rhythm of product hierarchy decisions about the priority of product
managers: How do we relate (unconstrained) sales to (constrained)
batches and the extent of postponing work-in-progress inventories
production? The response was to repair the S&OP sales forecast, as
till a time when it was clear what customers would order.
highlighted by the Manager of Manufacturing & Supply e ID,
M, D, and E decisions were required daily, as pointed out by the
Product Line Planning Manager A e ID, That is decided, yes, beyond 12 months, that forecasts need to be
unconstrained, but within 12 months, it may have to be con-
Daily planning! I would say the objective or the goal of the daily
strained if we have any constraints. The ideal is that in 12
planning (is) to maintain free availability. Free availability
months, we should be able to fix those constraints with our own
means you should have the right products on stock all the time.
manufacturing (and) with all possible supplies, and then be-
So we can serve the market … Also the daily (planning) means
tween 1 and 3 years, everything is available, and we should just
you should book the dispatch order every day, you should order
produce what they (sales) forecast.
material every day, yes, you have made your M decision on what
to produce, so we do this, we have a loop of tasks that we do
them each day … Here the core task for the supply chain Another reference, the adjusted S&OP sales forecast, which
manager is to daily or rather continuously maintain free avail- would only ‘un-constrain’ the forecast after 12 months, was un-
ability. Every second is important. We need to have free avail- derstood to be tolerable, so that certain excess customers would be
ability per product at every moment. re-considered after 12 months.
This heightened attention not only to ‘own manufacturing’ but
also to ‘all possible supplies’, produced another item of subsistence
In other words, objecting items of subsistence here ‒ the
in the form of an inter-factory production network. Factories with
product hierarchy, work-in-progress inventories, and product
excess capacity would increase their shipment to support other
batches ‒ were to be managed daily. The intersection between
factories with capacity constraints. The meeting minutes following

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx 11

Fig. 2. Presentation of S&OP and factory forecasts ‒ Pilot S&OP meeting June 2011 (green bars represent the monthly factory forecasts and red bars represent the monthly S&OP
sales forecasts; certain information is erased to preserve anonymity of the organisation.). (For interpretation of the references to colour in this figure legend, the reader is referred to
the Web version of this article.)

the above first PGP S&OP meeting provided a list of actions, for above, were ways to make factories’ planning systems flexible but
example, ‘(i)n Factory A,2 Channel xx to run 6 days for 24 hours’. this also attracted yet another objection from the demand chain. As
Product Line Planning Manager B e ID corroborated that this Fac- the Product Line Planning Manager B e ID explained, ‘factory A
tory A was indeed ‘supporting inventories to channels in other supported inventories to others, but the (adjusted) S&OP forecast
factories with capacity constraints’. did not include this shipment’. The S&OP sales forecast focused
The reference, the adjusted S&OP sales forecast, produced de- only on the end-customer in the market. This was why from May
cision activities to mitigate short-term capacity constraints. Excess 2011 factories struggled with product line planning because they
customers, while turned away then, could still be served later. The understood the adjusted S&OP forecasts as too low. Fig. 2 provides
tension between current and excess customers was addressed by a an illustration. It was one of the many PowerPoint slides shown in
forecast that created simultaneity of both daily product line plan- the June 2011 pilot S&OP meeting that visualised the difference
ning in factories, which concerned existing customers, and medium between unconstrained (adjusted) sales and constrained factory
to long term capacity planning in GDC and sales, which concerned forecasts. For most figures discussed in the meeting, the uncon-
excess (potential) customers. The factory network provided a way strained sales forecasts were lower than the constrained factory
to connect the two. forecasts because of their exclusion of shipments.
According to the Product Line Planning Manager A e ID,
5.7. Crossing point 7: internal customers More or less all factories now realise that their forecasts refer to
the shipment out of factories. During this period (February 2011
The implementation of the adjusted S&OP sales forecast to May 2011), there was a huge increase in stock in regional
impacted capacity. According to the minutes on the pilot S&OP warehouses for example the Singapore warehouse and a num-
meeting in May 2011, ber of factory warehouses. As a consequence, factories put sales
Both adjusted sales forecasts, around xxx euros, and factories forecast aside as they speak a different language. Sales always
forecasts, around yyy euros, were higher than the levels of consider the volume sold to customers, but the factories always
January 2011 because of the efforts put in place to increase shifts think in terms of shipment from either warehouses or factories
and manpower … Factory M was running all available hours (24/ to customers, and shipment from factories to factories.
7), started in Feb. 2011; Factory N was running overtime for all
channels on longer shifts; in Factory O, temporary extra night While sales anticipated total volume delivered to external cus-
shifts were added. tomers, factories planned for total volume of shipments out of
factories. For factories, the reference was the shipment volume and,
The factory network and ‘all possible suppliers’, as exemplified therefore, in addition to external customers, shipments within
inter-factory networks and to warehouses also counted. For sales,
the reference was only sales volume to external customers. Sud-
2
The actual locations of the factories are concealed for the purpose of anonymity. denly it was also clear why for most months the S&OP sales

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
12 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx

forecasts were lower than the factory forecasts. Product Line and reference. At the end of the fieldwork, it was not apparent to
Planning Manager B e ID pointed this out when describing the managers what new objections were attracted to the LTASB fore-
supporting work provided by Factory A, ‘forecasts (in Factory A) cast, but this did not guarantee their absences.
were higher than sales (forecasts). Sales simply didn’t take this into
consideration’. He also confirmed that there was ‘increase of
shipment from Factory B to the regional warehouses’, which sales 6. Accounting, simultaneity, and relative completeness in the
still ‘overlooked’. dynamic enactment of the demand chain
Also, the lead-time required to transport products between
factories and warehouses in different geographical regions 6.1. The enactment of the demand chain multiple
impacted forecasts, as explained by the S&OP PLP Manager ‒ GDC,
The empirical analysis is a detailed account of the development
In general, we have four geographical areas, North America, of forecasts as a moving engagement with subsistence of an object
South America, Europe and Asia. The rule is that there is no lead (Latour, 2007). The object, the demand chain, never shows itself in
time between the shipping factory and the ‘receiver’ if they are its entirety, but it is enacted as a multiple around coordination
in the same geographical area. When they are in different areas, problems created by intersections between forecasting mecha-
there is a 1e2 months lead-time. nisms and items of subsistence attracted to them. In this account, a
forecast never represents the market but helps enact multiple
The demand chain at this crossing point thus revealed items of differentiated practices linking products, production, capacity,
subsistence such as internal customers, including factories and customers, and market. The forecast provokes new subsistence
warehouses, and the lead-time used to transport products between because when it proposes a decision model for coordination, other
different continents. This tension between internal shipments and items of subsistence of the demand chain materialise as new ob-
external sales was translated into a lead-time adjusted shipment jections. Moving forecasts’ engagement with moving instantiations
based (LTASB) forecast. The Product Line Planning Manager B e ID of subsistence of the demand chain is summarised in Table 3.
gave a numerical example to illustrate how an existing (adjusted) Table 3 recounts the empirical story e accounting’s role in
S&OP sales forecast was transformed into a LTASB forecast enacting the demand chain e as a set of crossings with subsistence
considering the impact of the lead-time. (objections) and (established) references (decision models for co-
Fig. 3 illustrates that, for instance in the first column, if the sales ordination): An objection intersects with an established forecasting
forecast predicted the volume sold to external customers as 105 model, which presents a coordination tension and specifies a co-
and considering there was a lead-time of two months of shipment ordination concern to managers; each coordination concern breeds
from a European to an Asian warehouse, the European factory a new forecasting model; each model attracts and intersects with a
needed to ship six extra units today in order for the Asian ware- new objection from the demand chain, increasing its multiplicity.
house to have availability in two months’ time. The forecast This is a dynamic account of the enactments as discoveries and
therefore had to be 111. In columns where the LTASB forecasts was coordinations of the demand chain multiple.
lower than the S&OP sales forecasts, for instance, in the fourth Latour’s (2007) distinction between modes of reference and
month the adjusted S&OP sales forecast was 107 whilst the LTASB subsistence works differently in the case of the demand chain from
forecast was 106, the explanation from the product line planning his example of the horses in the museum. Firstly, the demand chain
manager was that the shipment forecast had to be reduced because may be more complex than horses in the field. The enactments of
some shipments were already forecasted two months earlier when the demand chain exist concurrently rather than in a history
(in the third month) the shipment forecast was increased to 118 separated by thousands of years. This shows that coordination
from the S&OP forecast (111). mechanisms never neatly substitute each other. This is why the
Challenging the adjusted S&OP sales forecast, inter-factory multiplicity of the demand chain is about simultaneity (as accu-
networks, internal customers, and the lead-time of transportation mulation) rather than a story that progressively unfolds as convo-
presented a tension whereby internal customers were not included luted, as is the case for horse evolution (Latour, 2007). There is
by sales. This tension raised another coordination concern in rela- continuity in the enactments of the demand chain multiple as
tion to what full capacity was. Here, a ‘shipment’ mode of fore- enduring coordination concerns, but this continuity is a process
casting for the demand chain was produced. This was not a that unfolds additional times and spaces rather than replaces
substitution of the adjusted S&OP sales forecast because the model existing ones. Secondly, subsistence exists both as activities and as
that allowed factories to submit a constrained forecast under ca- inscriptions. Subsistence is often made visible through inscription
pacity shortage was still in operation; so was the internal factory work (Robson, 1992; Robson & Bottausci, 2018); the world has to be
network. The additions were internal customers and lead-time made present, or presented, somehow, often by visual means
adjustments. References about the internal factory network, (Latour, 1999a). The difference between accounting and subsistence
existing and excess customers, warehouses, and lead-time co- is that accounting produces a transformation that organises sub-
existed. In this sense, the LTASB was another addition to the sistence. It adds to subsistence a decision mode that proposes how
adjusted S&OP sales forecast. subsistence be handled. Accounting transforms subsistence into
The empirical analysis ends here, but it does not mean that the decisions and actions upon subsistence. For each accounting, there
unfolding of the demand chain multiple is necessarily complete. As is an air of completeness because a decision model is juxtaposed
shown in the seven crossing points outlined here, adding a new next to a specified set of subsistence. Crossings between accounting
time and space requires a new crossing point between subsistence and subsistence set a condition for this, turning objecting subsis-
tence into a specific tension of the demand chain and a particular

Fig. 3. A numerical example of LTASB forecasts transformed from the adjusted S&OP sales forecasts.

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
 Table 3
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast

A summary of crossing points that enact the ‘demand chain’.

Crossing points Supply shortage Composition of Product and customer Average customers Planning Rhythms Capacity constraints Internal customers
the market variations

Subsistence e objections Empty inventories; lost Customers’ own Product volumes, Medium to long term Material families; Excess customers in Inter-factory network;
customers; complex voice; difference growth, novelty, value; production product hierarchy; the internal
delivery; between what customer heterogeneity and intermediary market customers; lead time
multiple customers said and machinery needs stocks of transporting
calculative agents did; general products across
market trend continents
Tension of coordination/related Customers not Particularity of Too many types of Investments plans Average customers at Current customers Internal customers not
coordination concerns served by operations customers not customers and too hampered by odds with daily product drive out included by
/How do we link compatible with many types of erratic customer line planning/How do we potential sales/What is full
supply to demand? generality of market products make requirements organise the planning customers./How do capacity?
trend/How do we link forecasting complex. /How do we relate rhythm in production? we
customers to the market? /How do we relate production relate sales to
customers to products? to sales? production?
References e coordination 36 months uncon 36 month market-based A, B, C, D, E products 36 months Factory forecast (daily); Constrained 12 Constrained 12

L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx

models strained market-based bottom-up forecast contributing 30%, 30%, unconstrained Order book; bill of months sales months
sales forecast 20%, 15% and 5% to market- materials; forecast; shipment forecast;
revenue (annual) based sales forecast capacity constraints unconstrained 12 unconstrained 12e36
based on e36 months market-based
minimised mean months market- shipment forecast
error from the based sales
previous 12 months forecast

13
14 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx

coordination concern (refer to the row of ‘Tension of coordination/ (Mouritsen, Hansen, & Hansen, 2009) with other accounting(s). The
related coordination concerns’ in Table 3), transforming coordina- success of one accounting does not depend on the failure of
tion tensions into decision models. another. This is related to the multiplicity of the demand chain,
This completeness is relative because the demand chain is wherein all enactments ‘become contemporary’ (Latour, 2005, p.
enacted as multiple managerial tensions. Accounting translates 30). Simultaneity happens because each crossing point develops a
subsistence into individual, and scattered, decision problems. The specific time‒space relationship between market, sales, factories,
general question of coordination is therefore delegated to a set of customers, products, and capacity, all of which co-exist in relation
only weakly coupled situations, each of which is internally to the demand chain. The generic market forecast e BCF e does not
coherent. Multiplicity happens because, while the scattered situa- fade away; it is present in both product line planning and central
tions may be coherent, the demand chain is not e it keeps planning. The adjusted S&OP forecast was not replaced by the
espousing new subsistence to leap in between situations. For LTASB forecast. Shipments were added to sales when there were an
instance, while serving the purpose of medium to long term ca- internal-factory network and regional warehouses.
pacity planning, the ‘smooth’ forecast was opposed by product line Simultaneity requires additions of decision mechanisms. This
planners because it did not consider the realistic customer. The helps coordination because each time‒space relation resolves a
‘smooth’ forecast creates comfort in one situation but turbulence particular tension as a forecast is mobilised as a source of handling
elsewhere. This dynamic is a consequence of accounting being both worrying subsistence. Each crossing point discovers a new objec-
performative and provocative. As a source of performativity it turns tion that presents another coordination tension and concern
each objection into a relevant tension and a specific decision model, somewhere else, which is then turned into an additional solution.
making the demand chain function in a certain way. As a force of Simultaneity has many orders, many ways of making the demand
provocation it helps new objections to emerge against what ac- chain work. A forecast helps to perform the discovery of emerging
counting reveals about the demand chain, adding new accounting subsistence because each produces concerns for others, and the
models to existing ones. This process may be ‘generative’ (Busco & success of coordination turns out to be the degree to which con-
Quattrone, 2015, 2018) but its resolution is hardly the result of cerns are tolerable.
managerial creativity and political compromise (Giovannoni & This adds to accounting research that sees accounting in-
Quattrone, 2018). Rather it is the accumulation of decision struments as providing coherence in action at a distance (Miller &
models that discover and organise coordination problems because Power, 2013; Robson, 1992). The study proposes that while one
of continuously emerging crossing points. tension may be organised by a centre of calculation, other streams
The finding that forecasts both add and organise coordination of activities that are difficult to bring into this centre are dealt with
problems suggests nuanced insights about forecasting and demand by other centres. There are actions on simultaneous various dis-
chain management generally (Vollman & Cordon, 1998; Vollman tances. Existing research also proposes that accountings may
et al., 2000) and S&OP processes specifically (Grimson & Pyke, compete in political games for influence (Edwards, Ezzamel,
2007; Lapide, 2005; Shapiro, 1998) in relation to lateral coordina- Robson, & Taylor, 1995; Mouritsen, 1999; Mouritsen et al., 2009;
tion of interdependent organisational functions (Atkinson, 2009; Preston et al., 1992), and then, as shown by this study, to a certain
Tohamy, 2008). It is important to maintain the general idea that the degree it is also possible that different accountings can be tolerable
market guides sales and production planning activities, but man- when they perform parallel activities to those of others (crossing
agers engage with many things that mediate relations between points). Therefore, it is ‘pertinent to trace continual changes in loci
production and market. These various realities enact their own of control’ (Quattrone & Hopper, 2005, p. 760) to emphasise not
relation with customers, production, products, capacity, and so on. only the changes in controls but also their multiplicity. While a
The translations shown in Table 3 illustrate that none of these re- forecast attracts objections, the accumulation of forecasts increases
alities exists passively. In effect there are multiple forecasts, each of the number of enactments of the demand chain. Objections are
which takes one particular concern into consideration. The effect of made tolerable through the simultaneity of accountings. This is
this delegation is both to deal with the present concern and often what simultaneity does: it performs many streams of decision
also to produce other concerns, as they draw attention to other activities.
subsistence. Therefore, the coordination problem is enduring, but
simultaneously complete forecasts enable managers to settle each 6.3. Accounting and relative completeness
situated tension relatively independently of others.
The second effort to make delegation tolerable is to create
6.2. Accounting and simultaneity completeness of accounting in each situated enactment of the ob-
ject. The demand chain multiple is realised by many crossing
It takes efforts to make delegation tolerable and operational. The points, each of which specifies a particular cumbersome subsis-
first effort is to understand time as simultaneity (Latour, 2005), tence, a manageable tension, a decision model and resources
instead of as succession. Simultaneity relativises distinctions such required for handling such a worrying situation. Completeness,
as short term and long term, and translates such temporal relativity therefore, translates subsistence into articulated decision proced-
into spatial co-existence. Various long terms, medium terms, and ures that handle subsistence; it juxtaposes reference and subsis-
short terms are enacted and differentiated along the demand chain tence. This completeness functions until another (absent)
as different tensions (refer to Table 3) that make the orientation of subsistence emerges and objects to an existing decision model. In
the time‒space coordinates of the forecast variable. The passage of this sense, accounting helps to complete the demand chain by
temporalities is therefore not primarily a matter of progression. providing each of its situated enactments with a decision
They are all contemporary and simultaneous in the enactment of mechanism.
the demand chain. This shows that lateral coordination, which the This sense of completeness is present in Corvellec et al. (2018),
forecast is expected to bring, is an effect of spatial and temporal who arrives at a similar conclusion when studying how accounting
work (Corvellec et al., 2018). invoices produce proximity between economy and environment.
As shown, simultaneity means that accountings accumulate and Such ‘distance’ work produced by accounting is complete since
cohabit. Each may function in its own setting and it neither sub- users at many levels from individual residents through to local
stitutes (Andon et al., 2007; Chua & Mahama, 2007) nor competes governments formed a coalition between the two otherwise

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx 15

opposite entities, money and environment. Money and environ- independently of each other. Yet each provokes objections, which
ment were made the same thing through an accounting invoice in turn makes present other absent subsistence and adds to the
that calculated the cost of not being environmental. Similar things multiplicity of the demand chain.
happened at each crossing point at EuroTech. Here, a relation is Secondly, there is a movement from accounting incompleteness
formed between subsistence and accounting, which are brought and instability to relative completeness; from contextualising one
closer to each other. Compared with Corvellec et al. (2018), the case incomplete accounting with complementing work to many com-
of EuroTech illustrates that delegation also breeds additional ob- plete accountings, each of which has a strong sense of formation
jections that point to yet other coordination concerns. As shown in and is an ‘engine’ of change. This happens because there is
the empirical analysis, complete accounting is relative in the sense emphasis on accounting for a situation, which requires a specific
that sealing off of one time‒space may provoke resistance from decision model. It is about the condition under which one is able to
other time‒spaces because any particular decision mechanism is rely on accounting. This completeness is also relative because it only
relevant only to a particular situation; it does not stop other ob- functions in a particular situation. It is an important mechanism to
jections. For instance, at crossing point 4, while sales and GDC were make multiplicity tolerable and acceptable. Tolerability is achieved
satisfied with their average customers, factories did not want to through accumulation of simultaneous accountings, and accept-
bear the burden of the average customer disrupting their daily line ability is realised via provision of relatively complete accountings.
product planning at crossing point 5. Coordination is, therefore, an Absence becomes material when it constructs specificity and
enduring process to discover tensions and to produce situated discontinuity. Specificity requires juxtaposition e proximity e be-
decision models. tween enacted realities and accounting (completeness) whilst
Relative completeness provides some nuances to insights discontinuity requires separability e organisation e of different
contributed by the literature on accounting incompleteness and enactments (simultaneity). This is neither a compromise nor a
instability (e.g., Ahrens & Chapman, 2004; Busco & Quattrone, sacrifice because different decision rules cohabit in their various
2015, 2018; Chua, 1995; Dambrin & Robson, 2011; Jørgensen & enactments of the object multiple.
Messner, 2010). This literature suggests complementing work Understanding accounting in performing simultaneity and
around accounting, for instance, strategising, institutions, politics, relative completeness has implications. When multiple calculations
dialogues, and so on, to make the incompleteness of accounting less create acentre of discretion (Andon et al., 2007; Quattrone &
straining. It is based on the premise that managers have, or develop, Hopper, 2001, 2005), simultaneity and relative completeness
a reservoir of extra knowledge about the absent realities they can ensure that multiplication and organisation of the ‘a’s are tolerable.
apply to contextualise incomplete and unstable accounting so that Here, acentre of discretion is translated into co-centres of actions.
accounting and other accounts, such as ‘strategic and operational Future research may consider further exploring the dynamic re-
arguments’ become ‘highly compatible’ (Jørgensen & Messner, lations between the performativity of accounting and multiplicity
2010, p. 188). The problems facing mangers at EuroTech are of objects because there may exist other relations between ac-
different. They insist that the forecast helps enact a world of counting, managerial actions, and enactments of an object
moving things and knowledge. Instead of relying on managers’ multiple.
sense making (e.g. Ahrens & Chapman, 2004, 2007; Jordan &
Messner, 2012; Jørgensen & Messner, 2010) or creativity (eg. Acknowledgments
Busco & Quattrone, 2015, 2018; Giovannoni & Quattrone, 2018) that
invites other things to help move accounting along, they propose We are grateful to managers and employees at EuroTech who
that accounting be actively bent towards objections and decisions; have generously shared their knowledge and contributed their time
they require that accounting keep adding objections and decision in participating in this research. We thank the editor, Professor
mechanisms upon objections; it is these additions that re-organise Chris Chapman, and the two anonymous reviewers for their
the many and variable relations that shape an object multiple. constructive comments on the previous versions of this paper. We
also thank Professor Paolo Quattrone, Professor Wai Fong Chua,
7. Conclusion Professor John Roberts, and Professor Paul Andon for their sug-
gestions on both earlier and current versions of the paper.
The case of EuroTech’s sales and operations planning (S&OP)
process explains how accounting, in the form of a forecasting
References
procedure, produces decision mechanisms, which in turn create
managerial attention to an object, the demand chain. Over time, Ahrens, T., & Chapman, C. S. (2004). Accounting for flexibility and efficiency: A field
accounting gains properties, partly because the multiplicity of the study of management control systems in a restaurant chain. Contemporary
object in question reveals new subsistence and partly because ac- Accounting Research, 271e301.
Ahrens, T., & Chapman, C. S. (2007). Management accounting as practice. Ac-
counting provokes new subsistence to appear. In this account, ac- counting, Organisations and Society, 32(1/2), 1e27.
counting plays an active part in the enactment of the demand chain Andon, P., Baxter, J., & Chua, W. F. (2007). Accounting change as relational drifting: A
multiple and its coordination. The study follows the inter-related field study of experiments with performance measurement. Management Ac-
counting Research, 18(2), 273e308.
movements of subsistence of the demand chain and of fore- Atkinson, W. (2009). S&OP: Now more than ever. Supply Chain Management Review,
casting, and it finds, firstly, that there is a movement from ac- 13, 50e54.
counting of to accounting for the object. The power of accounting Briers, M., & Chua, W. F. (2001). The role of actor-networks and boundary objects in
management accounting change: A field study of implementation of activity-
here is not merely an observation that accounting is constitutive, based costing. Accounting, Organisations and Society, 26(3), 237e269.
but how it becomes both performative and provocative; not only Burchell, S., Clubb, C., Hopwood, A. G., Huges, J., & Nahapiet, J. (1980). The roles of
that it makes territories visible, but how visibility becomes a accounting in organisations and society. Accounting, Organisations and Society,
5(1), 1e25.
resource for decision making and for discovery of new subsistence
Busco, C., & Quattrone, P. (2015). Exploring how the balanced scorecard engages and
that will require new decision models. This happens because of the unfolds: Articulating the visual power of accounting inscriptions. Contemporary
concerns with lateral coordination. There is a movement from time Accounting Research, 32(3), 1236e1262.
as succession to time as simultaneity. This happens because there is Busco, C., & Quattrone, P. (2018). In search of the “perfect one”: How accounting as a
maieutic machine sustains inventions through generative ‘in -tensions. Man-
accumulation of accountings, each of which helps enact its own agement Accounting Research, 39, 1e16.
reality of the demand chain. Each can operate more or less Busco, C., Quattrone, P., & Riccaboni, A. (2007). Management accounting: Issues in

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129
16 L. Yu, J. Mouritsen / Accounting, Organizations and Society xxx (xxxx) xxx

interpreting its nature and change. Management Accounting Research, 18(2), Merchant, K. A. (2006). Measuring general managers’ performances: Market, ac-
125e149. counting and combination-of-measures systems. Accounting, Auditing &
Chapman, C. S. (1997). Reflections on a contingent view of accounting. Accounting, Accountability Journal, 19(6), 893e917.
Organizations and Society, 22(2), 189e205. Merchant, K. A., & Manzoni, J. F. (1989). The achievability of budget targets in profit
Chua, W. F. (1995). Experts, networks and inscriptions in the fabrication of ac- centers: A field study. The Accounting Review, 64(3), 539e558.
counting images: A story of the representation of three public hospitals. Ac- Miller, P. B., & O’Leary, T. (2007). Mediating instruments and making markets:
counting, Organisations and Society, 20(2e3), 111e145. Capital budgeting, science and the economy. Accounting, Organisations and So-
Chua, W. F., & Mahama, H. (2007). The effect of network ties on accounting controls ciety, 32(7/8), 701e734.
in a supply alliance: Field study evidence. Contemporary Accounting Research, Miller, P. B., & Power, M. (2013). Accounting, organizing and economizing: Con-
24(1), 47e86. necting accounting research and organisation theory. The Academy of Man-
Corvellec, H., Ek, R., Zapata, P., & Zapata Campos, M. J. (2018). Acting on distances: A agement Annals, 7(1), 555e603.
topology of accounting inscriptions. Accounting, Organisations and Society, 67, Millo, Y., & MacKenzie, D. (2009). The usefulness of inaccurate models: Towards an
56e65. understanding of the emergence of financial risk management. Accounting,
Dambrin, C., & Robson, K. (2011). Tracing performance in the pharmaceutical in- Organisations and Society, 34(5), 638e653.
dustry: Ambivalence, opacity and the performativity of flawed measures. Ac- Mol, A. (2002). The body multiple: Ontology in medical practice. Durham: Duke
counting, Organisations and Society, 36(7), 428e455. University Press.
Edwards, P., Ezzamel, M., Robson, K., & Taylor, M. (1995). The development of local Mouritsen, J. (1999). The flexible firm: Strategies for a subcontractor’s management
management of schools: Budgets, accountability and educational impact. control. Accounting, Organisations and Society, 24(1), 31e55.
Financial Accountability and Management. https://doi.org/10.1111/j.1468- Mouritsen, J., Hansen, A., & Hansen, C.Ø. (2009). Short and long translations:
0408.1995.tb00234.x. Management accounting calculations and innovation management. Accounting,
Ezzamel, M., Robson, K., & Stapleton, P. (2012). The logics of budgeting: Theorization Organisations and Society, 34(6/7), 738e754.
and practice variation in the educational field. Accounting, Organisations and Petani, F. J., & Mengis, J. (2016). In search for lost space: The process of space
Society, 37(5), 281e303. planning through remembering and history. Organization, 23, 71e89.
Frandsen, A. (2009). From psoriasis to a number and back. Information and Orga- Preston, A. M., Cooper, D. J., & Coombs, R. W. (1992). Fabricating budgets: A study of
nisation, 19, 103e128. the production of management budgeting in the national health service. Ac-
Frow, N., Marginson, D., & Ogden, S. (2010). Continuous’ budgeting: Reconciling counting, Organisations and Society, 17(6), 561e593.
budget flexibility with budgetary control. Accounting, Organisations and Society, Quattrone, P. (2009). Books to be practiced: Memory, the power of the visual, and
35(4), 444e461. the success of accounting. Accounting, Organisations and Society, 34(1), 85e118.
Giovannoni, E., & Quattrone, P. (2018). The materiality of absence: Organizing and Quattrone, P. (2015). Value in the age of doubt: Accounting as a maieutic machine.
the case of the incomplete cathedral. Organization Studies, 39(7), 849e871. In M. Kornberger, L. Justesen, A. Madsen K., & J. Mouritsen (Eds.), Making things
Grimson, J. A., & Pyke, D. F. (2007). Sales and operations planning: An exploratory valuable (pp. 38e61). Oxford: Oxford University Press.
study and framework. International Journal of Logistics Management, 18(3), Quattrone, P. (2017). Embracing ambiguity in management controls and decision-
322e346. making processes: On how to design data visualisations to prompt wise
Hansen, A. (2012). Nonfinancial performance measures, externalities and target judgement. Accounting and Business Research, 47(5), 588e612.
setting: A comparative case study of resolutions through planning. Management Quattrone, P., & Hopper, T. (2001). What does organisational change mean? Spec-
Accounting Research, 21(1), 17e39. ulations on a taken for granted category. Management Accounting Research,
Hopwood, A. G. (1976). Editorial. Accounting, Organiations and Society, 1(1). 12(4), 403e435.
Hopwood, A. G. (1987). The archaeology of accounting systems. Accounting, Orga- Quattrone, P., & Hopper, T. (2005). A ‘timeespace odyssey’: Management control
nisations and Society, 12(3), 207e234. systems in two multinational organisations. Accounting, Organizations and So-
Jordan, S., & Messner, M. (2012). Enabling control and the problem of incomplete ciety, 30(7), 735e764. https://doi.org/10.1016/j.aos.2003.10.006.
performance indicators. Accounting, Organisations and Society, 37(8), 544e564. Quattrone, P., & Hopper, T. (2006). What is it? SAP, accounting, and visibility in a
Jørgensen, B., & Messner, M. (2010). Accounting and strategising: A case study from multinational organisation. Information and Organization, 16(3), 212e250.
new product development. Accounting, Organisations and Society, 35, 184e204. Revellino, S., & Mouritsen, J. (2015). Accounting as an engine: The performativity of
Knox, H., O’Doherty, D., Vurdubakis, T., & Westrup, C. (2015). Something happened: calculative practices and the dynamics of innovation. Management Accounting
Spectres of organization/disorganization at the airport. Human Relations, 68, Research, 28, 31e49.
1001e1020. Robson, K. (1992). Accounting numbers as ‘inscriptions’: Action at a distance and
de Laet, M., & Mol, A. (2000). The Zimbabwe bush pump: Mechanics of a fluid the development of accounting. Accounting, Organisations and Society, 17(7),
technology. Social Studies of Science, 30(2), 225e263. 685e708.
Lapide, L. (2005). An S&OP maturity model. Journal of Business Forecasting, 24(3), Robson, K., & Bottausci, C. (2018). The sociology of translation and accounting in-
15e20. scriptions: Reflections on Latour and accounting research. Critical Perspectives
Latour, B. (1987). Science in action. How to follow scientists and engineers through on Accounting, 54, 60e75.
society. Cambridge, MA: Harvard University Press. Shapiro, J. (1998). Bottom-up versus top-down approaches to supply chain
Latour, B. (1999a). Circulating reference: Sampling the soil in the Amazon forest. In modelling. In Quantitiative models for supply chain management (pp. 739e759).
B. Latour (Ed.), Pandorra’s hope: Essays on the reality of science studies (pp. Dordrecht: Kluwer Academic Publishers.
24e79). Boston: Harvard University Press. Skærbæk, P., & Tryggestad, K. (2010). The role of accounting devices in performing
Latour, B. (1999b). When things strike back: A possible contribution of ’science corporate strategy. Accounting, Organisations and Society, 35(1), 108e124.
studies’ to the social sciences. British Journal of Sociology, 51(1), 105e123. Tohamy, N. (2008). The evolution of S&OP. Supply Chain Management Review, 12,
Latour, B. (2005). From realpolitik to dingpolitik. In B. Latour, & P. Weibel (Eds.), 10e11.
Making things public (pp. 14e44). Cambridge: MIT Press. Vaivio, J. (1999). Examining ‘the quantified customer’. Accounting, Organisations and
Latour, B. (2007). A textbook case revisited: Knowledge as mode of existence. In Society, 24(8), 689e715.
E. J. Hackett, M. Lynch, J. Wajcman, & O. Amsterdamska (Eds.), The handbook of Vollman, T. E., & Cordon, C. (1998). Building successful customer-supplier alliances.
science and technology studies (pp. 83e112). Cambridge: MIT Press. Long Range Planning, 31, 684e695.
Law, J., & Singleton, V. (2005). Object lessons. Organisation, 12(3), 331e355. Vollman, T. E., Cordon, C., & Heikkila, J. (2000). Teaching supply chain management
Marginson, D., & Ogden, S. (2005). Coping with ambiguity through the budget: The to business executives. Production and Operations Management Journal, 9,
positive effects of budgetary targets on managers’ budgeting behaviours. Ac- 81e90.
counting, Organisations and Society, 30(5), 435e456.

Please cite this article as: Yu, L., & Mouritsen, J., Accounting, simultaneity and relative completeness: The sales and operations planning forecast
and the enactment of the ‘demand chain’, Accounting, Organizations and Society, https://doi.org/10.1016/j.aos.2020.101129