Why the HEART framework?
• IIAAYKIG - If it’s an acronym, you know it’s good
• Metrics are tougher than you think!
• It’s a simple way to ensure you’re thinking about every aspect of the
user journey / way a user sees your product
for X
Uber + Alcohol + Lemonade
UBER FOR
SPIKED
LEMONADE
Photo credit: [Link]
Happiness
Engagement
Adoption
Retention
Task Success
Happiness - How happy is your user?
Engagement - How engaged is your user in the short term?
Adoption - How many interested users have actually tried your product?
Retention - How many users do you retain long term?
Task Success - How successful are you at allowing users to perform the most valuable task?
Goals Signals Metrics
What is the thing we need to A signal expressed over time.
What do you want to happen?
measure? This is the thing you watch.
Goals Signals Metrics
Happiness
Engagement
Adoption
Retention
Task Success
1. Happiness
2. Engagement
3. Adoption
4. Retention
5. Task Success
1. Adoption
2. Task Success
3. Engagement
4. Retention
5. Happiness
Goals
• Maximize drinker satisfaction with our
Happiness spiked lemonade and the delivery
service
• Maximize # of orders and total value of
the orders (through add-ons like
Engagement additional shots in the lemonade or
extra fast delivery)
• Maximize the # of people who order at
Adoption least 1 glass of spiked lemonade
• Maximize the % of users that return
Retention and place a repeat order on a monthly
basis
• Minimize # of "abandoned
carts" (people that don't complete
purchase)
Task Success • Maximize the # of users that
successfully complete an order within
5 minutes
Signals
• App store rating
Happiness • Scores from the NPS survey sent to
customers
• # of orders per customer
Engagement • Order value in dollars per customer
• # of users who order at least 1 glass of
spiked lemonade
Adoption • # of users that have downloaded and
opened the app
Retention • People that order spiked lemonade
• Incomplete orders
Task Success • Time taken to place each order, from app
start to "success" screen
Metrics
• App store rating change, month over month
Happiness • % of perfect 10 NPS scores
• Average order value in dollars per day
Engagement • Average # of orders per day per user
• % of new users -> customers
(# of new users that order divided by the # of
Adoption people that have opened the app to get a
percentage)
• Retained users
(# of people ordering this month divided by
Retention the # of those same people that ordered last
month to get a percentage)
• Average time to checkout
Task Success • % of orders incomplete per month
Goals Signals Metrics
• Maximize drinker satisfaction with our • App store rating
• App store rating change, month over month
Happiness spiked lemonade and the delivery • Scores from the NPS survey sent to
• % of perfect 10 NPS scores
service customers
• Maximize # of orders and total value of
the orders (through add-ons like • # of orders per customer • Average order value in dollars per day
Engagement additional shots in the lemonade or • Order value in dollars per customer • Average # of orders per day per user
extra fast delivery)
• # of users who order at least 1 glass of • % of new users -> customers
• Maximize the # of people who order at spiked lemonade (# of new users that order divided by the # of
Adoption least 1 glass of spiked lemonade • # of users that have downloaded and people that have opened the app to get a
opened the app percentage)
• Retained users
• Maximize the % of users that return
(# of people ordering this month divided by
Retention and place a repeat order on a monthly • People that order spiked lemonade
the # of those same people that ordered last
basis
month to get a percentage)
• Minimize # of "abandoned
carts" (people that don't complete
• Incomplete orders
purchase) • Average time to checkout
Task Success • Maximize the # of users that
• Time taken to place each order, from app
• % of orders incomplete per month
start to "success" screen
successfully complete an order within
5 minutes
Tips on using HEART
• Flexible – use it for a product area or a whole company
• Use it for reporting, not exploring
• It’s not strict and you don’t have to use them all – try to pick one that matters
most
• You can use the signals column to inform engineers of tracking requirements