More Related Content
Similar to Building a business with data (20)
Building a business with data
- 1. Data maturity Building a business with data 1
- 2. uSwitch was part of a £2.2bn acquisition of ZPG
- 3. After seeing steady growth and 4 acquisitions
- 4. The approach to data influenced many areas ◉ Profitability ◉ Competitive position ◉ Business model ◉ Team structure It evolved as we grew 4
- 5. 3 stages of data maturity as we grew 5 1Control Data Past 2Flexibility Information Present 3Influence Insight Future
- 6. Understanding The business model and challenges 6 0
- 7. uSwitch is a 2 sided platform Market aggregation Access to volume Free to use Fee per transaction Consumers benefit from more sellers Sellers benefit from more consumers
- 8. Four key areas of optimisation Marketing Product Commercial Regulation
- 9. Performance under previous ownership
- 10. Due diligence summarised the issues ◉ Inability to respond to competition ◉ Poor supplier relationships ◉ Overstretched teams 10
- 11. Root cause was ability to manage data ◉ Lack of data expertise ◉ Little visibility of performance ◉ Common tasks with data broke systems 11
- 12. Control Learning to count correctly 12 1
- 13. Goals and outcome ◉ Control how data is managed ◉ Understand what has happened ◉ Measure impact of changes Improve profitability
- 14. ◉ An authoritative copy of all crucial data ◉ Robust data provenance and governance controls ◉ A common language across business units ◉ Self service A single source of truth (SSOT)
- 15. ◉ Describes how to think with data ◉ See the world more clearly and accurately ◉ Better data guided decisions Statistical thinking
- 16. 16 Letting humans be humans Photo by Rock'n Roll Monkey on Unsplash
- 17. A SSOT helped change what question we could answer ◉ “I have spent £x and made £y” ◉ “I understand the funnel and am able to see problems quickly” ◉ “This happened when we changed this system” Graduate roles replaced by analysts
- 18. 18 Doubling conversion Photo by rawpixel on Unsplash
- 19. Randomised controlled trials ◉ Customers to the platform are randomly assigned into groups ◉ An experimental group who see the new part of the site being tested ◉ A control group sees what was there before ◉ Difference in response between the groups is measured statistically
- 20. Example of an A/B testing dashboard
- 21. Conversion rate improvements across uSwitch
- 22. 22 Finding £1,000,000 Photo by Sharon McCutcheon on Unsplash
- 23. Example email from energy supplier
- 24. Lessons from getting data under control ◉ Changes the types of questions you can answer ◉ Improvements across processes compound ◉ Don’t overestimate partners capability with data ◉ Data can assist most decisions
- 25. Flexibility Creating information from data 25 2
- 26. “ “Information is data endowed with relevance and purpose. Converting data into information thus requires knowledge. And knowledge, by definition, is specialized.” Peter Drucker.
- 27. Goals and outcome ◉ Allow multiple versions of the truth ◉ Understand what is happening ◉ Deeper understanding of the market Improve competitive position
- 28. Multiple versions of the truth
- 29. 29 Becoming market leader Photo by Steven Lelham on Unsplash
- 31. Google data on paid search performance
- 32. 32 The “Ewan shuffle” Photo by Amanda C on Unsplash
- 34. 34 Meta products Photo by Coline Beulin on Unsplash
- 35. Automated P&L
- 37. Lessons from being flexible with data ◉ Data flexibility requires a change in mindset ◉ Meta products continue to optimisation ◉ Data influences how you are structured
- 38. Influence Creating and sharing insight 38 3
- 39. Goals and outcome ◉ Influencing external decisions ◉ Predicting what will happen ◉ Innovation and new revenue streams Protecting and evolving the business model
- 40. ◉ Need to manage more data and answer harder questions ◉ Combination of algorithms, technology and scientific methods to extract insights ◉ Employs techniques and theories drawn from many fields Data science
- 41. 41 Energy inSight Photo by Coline Beulin on Unsplash
- 42. Energy supplier inSights dashboard
- 43. Impact ◉ British Gas changed hedging strategy ◉ Suppliers teams have been replaced by our tools ◉ MDs are regular users of the product ◉ We hired the first user from British Gas Created a dependency on our platform and data
- 44. 44 Influencing the regulator Photo by Hunters Race on Unsplash
- 46. Impact ◉ Influence those decision by more objective measures ◉ It established a strong relationship with Ofgem ◉ Enabled us to understand the impact of these changes and to plan early
- 47. 47 Predicting the future Photo by Hunters Race on Unsplash
- 48. We can predict the impact of every plan
- 49. Choosing who to partner with
- 50. Lessons from influencing with data ◉ Data helped with each sale ◉ Capture everything ◉ Data needs a story
- 51. Whats next? 51 ?
- 52. Each new technology lowers the cost of something Used in novel ways 52
- 53. AI and machine learning are making prediction cheap Everything will be customised around ourselves 53
- 54. Conclusion ◉ Data has had a huge impact on uSwitch’s success ◉ A data assisted approach accelerated business growth ◉ All businesses need to adapt as new technology emerges Data leadership needs to start at the top
- 55. Any questions ? You can find me at ◉ [email protected] Thanks! 55