Retention + Engagement

Question 1

Retention

Answer 1

How many users remained active within a defined time period after signing up?

Question 2

Retention Knowledge Map

Answer 2

1. Retention: Retention refers to the amount of users still active within a time period and the length of time they are retained for. All activation, engagement, and resurrection efforts tie back to increasing retention.

1.1 Activated, A user who has established a habit around the core value prop.
1.1.1 Setup Moment, the user has done actions to setup for the core value prop.
1.1.2 Aha Moment, The aha moment is when the user experiences the core value for the first time
1.1.3 Habit Moment, The habit moment is when the user first signals they have established a habit around the core value prop
1.1.4 Activation, The process of taking a user from signup to the habit moment, establishing a habit around the core value prop.
1.1.5 Non-activated, A user who has not reached the habit moment.

1.2 Engaged, A user who has activated, is using the product within the natural frequency, Engagement is a spectrum of depth of usage
1.2.1 Power Users, the segment of users with the highest engagement
1.2.2 Core Users, the segment of users with medium engagement
1.2.3 Casual Users, the segment of users with the lowest engagement
1.2.4 At Risk, The direction of engagement is decreasing signaling a higher probability of becoming dormant.

1.3 Dormant, A user who has activated, been engaged, and then become unengaged.
1.3.1 Voluntary Dormant User, A user who chooses to stop using the product due to reasons like price, other options, missing features, or interest.
1.3.2 Involuntary Dormant User, A user who becomes dormant for reasons other than explicitly choosing to not use the product such as expired credit card, switching jobs, business closed, etc.
1.3.3 Resurrection, The process of taking a user from the dormant state back to the retained state.
1.3.4 Churned, A user who has activated, become unengaged, and been in a dormant phase for a period of time where the probability of resurrecting is extremely low.
1.3.5 “Hail Mary” Resurrection, The process of taking a user from a churned state back to a retained state

Question 3

Retention - Discover The Natural Behavior Use Cases

Answer 3

What is the natural usage behavior and pattern of my target audience?

Nature vs Nurture
Internal vs external types of triggers
Nature: The inherent behaviours and natural usage patterns of our target audience around our core value prop.
Nurture: What we manufacture to amplify and nurture those natural behaviours.

Defining your use case
1. Problem
2. Persona
3. Why
4. Alternative, not competitors
5. Frequency

Daily, Weely and Monthly are in the habit zone
Yearly and Years+ are in the forgettable zone

Question 4

Goldilocks Situation

Answer 4

Used to describe a situation in which something is or has to be exactly right

Question 5

Retention - Define Retention Metric

Answer 5

1. Frequency, using frequency histogram. See where the majority of the distribution is around.
2. Core behaviour
   2.1 Form groups that successfully did that action for successive periods
   2.2 Create a cohort chart for different action hypotheses
   2.3 Analyze the retention by comparing. We are looking for the action that leads to the flattest + highest retention.
3. Who

Common Mistakes
1. Combine actions
2. Using revenue metrics

Question 6

Retention - Visualizing Retention

Answer 6

1. Liftcycle Bar Chart, Visualizes the flow in and out of various user states within a certain time period.
   1.1 Quick Ratio = (New + Resurrected) / Loss, for example, when the quick ratio is 4, it means for every 4 active users we added, we lost 1.
2. Cohort Chart, Heat map visualization of average or individual cohorts showing “hot” or “dead” spots
   2.1 Absolute
   2.2 Percentage
   2.3 Absolute relative
   2.4 Percentage relative
   2.5 Avg relative, how is the cohort performing relative to the average?
3. Retention Curve, line graph visualization of the average or individual cohorts showing the shape of your retention over time.

Question 7

Retention - Analyzing Retention

Answer 7

1. High-level Diagnosis, Do you have a Bad (Trend to 0), Ok (Flatish), Good (Flat) and Great (Smile shape) business?
   When retention line is flat, that rate is the retention rate.
   1.1 What is the retention rate we should aim for?
   1.1.1 Social Apps (around 60%), Productivity (Around 35% - 40%), Sub E-Comm (Around 25% - 40%), Enterprise (8-% - 90%), Mid Market (Around 75%) and VSB (50% - 60%)
   1.1.2 Thinking about RR from first principles, what retention rate do I need to build the type of business I want?
   1.1.2.1 5 Year Retention
   1.1.2.2 Estimate 1 Year Value
   1.1.2.3 Total Addressable Market TAM Scenario, 100% market -> 10%, and different rates of RR
2. Individual Cohort Trends, How is my retention changing over time? And where does the shift happen? Outliers and Why? Look for Diagonal Stripes, we will see there is a diagonal stripe of red meaning that all users are affected by an event at the same time (outages, bugs, pricing, competitions, etc)
3. Segmentations, What are potential areas to improve?
   3.1 Persona
   3.2 Acquisition Params
   3.3 Device/Permissions
   3.4 Geography/Demo
   3.5 Product Categories
   3.6 Feature
   3.7 Did X In Y time
   3.8 Did X Y times In Z

Question 8

Activation

Answer 8

The process of taking a user from signup to establishing the habit. There are three primary moments.
1. Setup moment
2. Aha moment
3. Habit moment

Activation is typically highest highest-impact area

Question 9

Defining Our Activation Flow

Answer 9

1. Success, Start by revisiting the destination we want to get users to
2. Habit, Define the moment and metric that they have established a habit.
3. Aha, Define the moment and metric that they have experienced the core value prop.
4. Setup, Define the moment and metric that sets them up for the aha moment.
5. Analyze, Break our activation flow down piece by piece to generate hypotheses.

Question 10

Defining Habit Moment

Answer 10

Work backwards from the engagement/retention metric

The user has established a habit around the core value proposition

Qualitative: Problem + Frequency + Core Action. The problem indicates how high of a effort, the natural frequency indicates how often, and the core action in the retention metric indicates the behaviour

XaY
The number of times X the user has done the core action a within the initial time period Y

Verifying Habit Moment
1. Exploration, Explore the data by segmenting cohorts or retention curves to form some hypotheses
2. Correlation, Run an analysis to understand correlation with long term retained users. Build a Habit Moment Matrix
2.1 Y axis is core action count, and the X axis is Within the time period
2.2 Each cell contains Correlation (Tells us how strong the relationship with retention, closer to 1 the better), Sample Size (How many people took this many actions within this time period? The larger the better), Positive Predictive Value (Tells us the percentage of people that took this action that ended up being retained), Negative Predictive Value (Tells us the percentage of people that didn’t do this action that didn’t end up being retained
2.3 Remove any with correlation < .3, Remove those with small sample size. Look for inflection points. Inflection points in correlation are huge hints towards your metic. The point suddenly increases dramatically.
3. Causation, Run experiments to establish causation between the habit moment and long-term retention.

Question 11

Defining Habit Moment - How Long To Build A Habit?

Answer 11

It depends on the behaviour, people situation, etc.

Question 12

Defining Aha Moment

Answer 12

The user has experienced the core value prop for the first time. Intuitively when the Aha moment happens it feels like you have gained a special ability you didn’t have before.

fXaY
The first # of times X the user has done the core action a within the initial time period Y
Y need to be < then natural frequency

Verifying Aha Moment
Similar to the habit moment
Build a Aha Moment Matrix
Y axis is the core action(s), and the X axis is compared with the habit moment and long-term retention. Start with the core action + natural frequency. Then layer on more measurable actions
Within the cell, it measures the same thing.

Question 13

Defining Setup Moment

Answer 13

The user has done the actions to set up the core value prop.
What is the must-have information you need to deliver the aha moment?

XaY
The number of times X the user has done the setup action a within the initial time period Y

Build a setupMoment Matrix
Y axis is the core action(s), and the X axis is compared with the aha moment, habit moment and long-term retention. Similar to the other moments you should also look at variations of the time period + number of actions

Question 14

Analyzing Activation Flow

Answer 14

1. Qualitative comparison, What do successful users experience that unsuccessful users don’t between each moment?
2. Survival Analysis, How do different cohorts perform through the activation flow?
   2.1 Do certain cohorts perform very differently vs average? Why?
   2.2 What is the trend for each moment over time?
   2.3 Where are we losing the most users?
3. Survival Segmentation, What are other factors/characteristics that impact performance in activation?
   3.1 Same 3 questions as above

Question 15

Activation Metric Analysis

Answer 15

4 Types from easy to hard
1. Retention Curve Segmentation
Pros: Typically easiest/quickest analysis to do, great for exploring hypotheses
Cons: Lacks some accuracy and rigor. Easy to miss things like cohort/sample size
2. Venn Diagram Analysis
Pros: Easy to communicate verbally and visually, Good balance between sample size, ppv and npv.
Cons: May not be as accurate as regression models. Narrow view on hypotheses
3. Correlation Analysis
Pros: A concept most people are familiar with.
Cons: Harder to communicate visually. Narrow view on hypotheses.
4. Other Statistical Models
Pros: Great for wide exploration of hypotheses with statistical rigor. Can be used to build predictive scoring models
Cons: Required significant help from data scientist/analyst. Harder to interpret. Co-linearity (two hypotheses are correlated with each other)

Question 16

Activation Metric Analysis - Setting Up Data

Answer 16

1. Define Success, What is the success outcome we are trying to predict?
   1.1 We need to pick a spot on the retention curve as success, e.g. the user is still retained 3 months later
   1.1.1 Choose a spot on the flat part of the curve
   1.1.2 Not too close to the habit period not too far
   1.1.3 General guideline, for monthly, pick 6-9 months, for weekly, pick 3 months, for daily, pick 28 days or 1 month.
2. Binary Classification On Success, Did the user achieve our success definition or not?
3. Setup Hypotheses, What are the habit metric hypotheses we want to test?
4. Binary Classification On Hypotheses, Did the user complete the action hypothesis or not?

Question 17

Activation Metric Analysis - Venn Diagram Analysis

Answer 17

1. Build Venn Diagram Matrix, Build the VD analysis matrix with your hypotheses
   1.1 Draw a circle with all retained users
   1.2 Draw circles for each hypothesis, and see their overlaps.
   1.3 Maximize overlap, minimize outsiders and with meaningful size in the overlap.
   1.4 For the matrix, on Y axis is hypothsis with different number of actions, on X axis, they are Did Action, Did Action & Retained and Did Not Do Action and Retained
2. Calculate Venn Diagram Score. Calculate the VD score.
   Score = Did Action & Retained / Did Not Do Action and Retained + Did Action
3. Analyze Results, Analyze the scores and results to narrow in on a metric. We are looking for score that is > 50% which means highly predictive
   The 3 typical patterns
   3.1 Flat + Low %, The spectrum of scores remains flat as you increase the number of actions and the scores are all relatively low. This means that that action is not predictive even as the user does more of the action.
   3.2 Flat + High %, The spectrum of scores remains flat as you increase the number of actions and the scores are all relatively high. This means that that action is predictive but the more a user does the action does not influence the probability they will retain.
   3.3 Peak, The scores increase, peaking, then decrease. This means that as you increase the number of actions it becomes more predictive to a point, but at some point it decreases likely because the sample size is getting very small.
4. Repeat for Aha + Setup Metric. Repeat the same steps for aha and setup metric. Similar to how to define those metric, we work backwards, for Aha metric, we use Habit metric as the success metric.

Question 18

Activation Metric Analysis - Correlation Analysis

Answer 18

1. Build Matrix, Which hypotheses do I want to evaluate? Similar to the step in VD
2. Calculate Correlation, What is the correlation with long-term retention?
   R = Covariance (Retained, Hypothesis) / Std Dev (Retained) * Std Dev (Hypothesis)
   Look for anything > 50%
3. Calculate NPV + PPV, What is the NPV and PPV of this hypothesis?
   PPV: The % probability that if a user does the action they will retain
   PPV = Did Action & Retained / Did Action & Retained + Did Action & Did Not Retain
   NPV: The % probability that if a user does not do the action they will not retain
   NPV = Did Not Do Action & Did Not Retain / Did Not Do Action & Did Not Retain + Did Not Do Action & Retained
4. Calculate Sample, What is the total number of people who took this action?
5. Analyze Results, What is the metric that stands out as best? Correlation close to 1. Larger sample size. PPV and NPV are close to 100%

Question 19

Activation Strategies - The Four Fits

Answer 19

The most common reason a user doesn’t activate is if the activation flow doesn’t match the context at sign-up.

4 things users ask and we should also ask ourselves
For users <-> For You
1. Is this for me? <-> Who are they? ==> Audience Fit
2. Does this do what I want? <-> What do they want? ==> Promise Fit
3. How much do I care about solving this right now? <-> How badly do they want it? ==> Intent Fit
4. Do I know how to get it? <-> Do they know how to get it? ==> Knowledge Fit

Question 20

Activation Strategies - The Four Fits - Audience Fit

Answer 20

Users are looking for signals and affirmations that the product is for them throughout the whole activation flow.

1. Language, are you using the vocabulary, tone, and style they use?
2. Visual, do the visual elements align with who they are?

Question 21

Activation Strategies - The Four Fits - Promise Fit

Answer 21

We need to show them how they are getting closer to the promise along the way

Question 22

Activation Strategies - The Four Fits - Intent Fit

Answer 22

Candy vs Vitamin vs Painkiller

Question 23

Activation Strategies - The Four Fits - Knowledge Fit

Answer 23

How knowledgeable is the user about your experience?
Good combination is
Low Forefulness+ High User Knowledge
High Forefulness+ Low User Knowledge

Question 24

Activation Strategies - The Four Fits - Putting The Four Fits Into Action

Answer 24

1. Profiles, assess this on all the four fits using below scale
   1.1 Singular
   1.2 Clear Buckets (Segmentation)
   1.3 Gradient, wideband (Personalization)
2. Signals, predict which profile they are in
   2.1 Acquisition Parameters
   2.2 Supplemental Data
   2.3 User Asks
3. Segmentation, clear different paths for each bucket
4. Personalization, personalize the experience for each user.

Question 25

Product + Notifications + Incentives + People (PNIP)

Answer 25

Product: Everything starts with the product experience of activation.
Notifications: Emails, push, SMS, etc. The notification strategy must support the core product experience.
Incentives: Incentives can be used to accelerate/lubricate the product and notification experience, but are not a stand-alone strategy. Loyalty, discount, status, etc.
People: Customer Support, Customer Success. The use of people must support the Product, Notification, and Incentive experience.

Impact from high to low
For B2C, Product > Notification > Incentive > People
For B2B Mid Product > People > Notification > Incentives
For B2B SMB Product > Notification > People > Incentives

Product experience is the foundation, the rest is optimization.

Question 26

PNIP - Product

Answer 26

How to think through the product layer?

1. Do, The version of the product experience you give the user to do the action
   Full Product Experience <—> Slimmed Product <—> Customized
   Low forcefulness <—————> High forcefulness
2. Show, The visual elements you use to show them where to do the action
   Low forcefulness <—————> High forcefulness
   Pointer (highlights instruction) <—> Pulses (highlights the actual action) <—> Blackouts <—> Foced Action
3. Tell, The language you use telling them what to do
   Low forcefulness <—————> High forcefulness
   Guide <—> Suggest <—> Exact Entry
4. Motivational Boosts, The visual and language you use to add motivation to complete the action.
   Low forcefulness <—————> High forcefulness
   Badges <—> Social Proof <—> Progress Bars <—> Checklists <—> Countdowns <—> Scarcity

Three Principles Product Layer
1. Order of Operations: Do -> Show -> Tell -> Motivation. As Do has the most impact and each layer is building on top of another layer.
2. There are Pros/Cons to the level of forcefulness.
Low Forcefulness
Pros: Teaches user how to do something within full experience. Less false positives.
Cons: Lower conversion on near-term action.
High Forcefulness
Pros: Higher conversion of that short-term action. Good for solving cold starts and one-time information fathering.
Cons: Users tend to not learn how to do things within the context of the full experience.
3. Apply to all steps of the activation experience

Question 27

PNIP - Notification

Answer 27

8 channels for engagement
1. Email
2. Mobile
3. In Product
4. Platform, Facebook, Twitter, Slack
5. Browser
6. Desktop
7. Paid Media
8. Direct Mail

Don’t think like a marketer. Take off your marketing hat, and think like a personal assistant.

Four Layers of Notifications
1. Action. The action you want them to complete in the product in response to the notification
2. Destination, The product experience of where they land in the product in response to the notification
3. Hook, The message and hook that gets a user to response
3.1 The key mistake is trying to teach a behavior within the notification itself that is out of context.
4. Trigger, the action or inaction plus the time of when to send the notification.
The key mistake is sending a notification while the user is still in the product.
2nd key mistake is time-only based triggers vs action + who + time
3rd key mistake is failing to understand the downside when determining success, unsubscribe for example

Aim for high Specificity across 4 layers

Question 28

PNIP - Incentives

Answer 28

3 Types
1. $$$$
2. Time, for example, timed trials
3. Status, Badges, elite programs, etc

Question 29

PNIP - People

Answer 29

People strategies should support product strategies

3 Ways to Use People Strategies
1. Research, use it as a research function to improve the activation experience
2. Save, Saving those who had a bad experience
3. Low Knowledge Fit, When complexity is high, or knowledge is low but the customer is otherwise a good fit.

Question 30

Perverse Incentives

Answer 30

Perverse incentives are incentives that result in unintended negative consequences due to actions people take to receive the incentive. Perverse incentives are a common occurrence because decision-makers don’t always think things through before acting, often with disastrous results.

Question 31

Creating Habit Moment Experience

Answer 31

The Basic Building Habits
1. Trigger, the cue that inspires action
1.1 Organic Triggers
1.2 Manufactured Triggers
2. Action, The core action you want the user to take in response to the trigger
3. Reward, the reward the user gets in response to taking the action. Remind them of the reward to increase effectiveness.

The goal is to get the user to establish the organic habit loop

4 Strategies to Accelerate Habit Creation
1. Manufactured Loops
1.1 Time trigger, JetBlue: Your flight to LAX in 90 mins
1.2 Location trigger, Starbucks: Pay with Starbucks app when near favorite store
1.3 Change trigger, Gametime: SF Giants tickets price just decreased to $20
1.4 Peer trigger, Medium: John just posted 10 ways to learn faster
1.5 Programmatic trigger, Dropbox: Backup your photos when connect phone to computer.
2. Environment Loops
2.1 Build
2.2 Buy, buy access via ads
2.3 Partner, if closed platform, a way to partner?
3. Use Case Transition
3.1 What is a more frequent organic trigger related to the problem/need?
4. Grand Exit, Increasing delta (10X better) of aha moment experience. The grand exit tends to not be a sustainable strategy over the long term.

Question 32

Creating Aha Moment Experience

Answer 32

4 Pieces To The Aha Product Experience
1. Core Action
2. Warm Start. The experience that creates momentum toward the core action
3. Supporting Actions
4. Empty State. The empty states of other features should lead back to the core action. E.G. empty dashboard.

Question 33

Creating Setup Moment Experience

Answer 33

3 Types of Must-Haves
1. User Info
2. Permissions
3. Social, To get value, do they need others in the experience?

Notification at this moment is just a distraction. The strategy should be dropout detection and immediate recovery.

Question 34

Evaluating Your Activation Experience End To End

Answer 34

Worked backward, and evaluated forwards. Then repeat.

Use the Psych framework to evaluate the energy

Negative Psych
1. Physical, Elements that require the user to do something (filling out a form)
2. Cognitive, Decisions that the user needs to make or questions they need to figure out.
3. Fit Mis-Alignment, Elements that misalign with Audience Fit, Promise-Fit, or Intent Fit.

Positive Psych
1. Fit Alignment
2. Rewards
3. Motivational Boosts, “Carrots” that lead the user to the value. They provide the least amount of positive Psych and won’t work unless you have Fit Alignment and show genuine rewards.

Starting Psych is influenced by the four fits
Reordering matters

Question 35

The Four Strategies To Deepen Engagement

Answer 35

1. Adding Use Cases, Moving users to adopt additional use cases of the product. Usage is driven by the use case and natural frequency. Users don’t enter with the mindset of all use cases. You need to layer them over time. Really expensive to develop.
2. Increase Frequency, Increase the frequency of use of the product. Might become a reason to quit.
3. Increase Feature Usage. Increasing the % of features that users are using within the product. Within a use case, there is a surface area of features. More unused features might create a mental tax. A feeling of not getting the full worth
4. Increase Intensity, Increasing the intensity of each use of the product.

Question 36

Measurement Of Your Engagement Strategy

Answer 36

1. Total Engagement, what is the total engagement of my active user base?
2. Engagement Per Active, what is the average engagement per active? High-level trends.
3. Engagement Segmentations. We want to have strategies to move users from casual to core and then to power while creating business value.

Question 37

How to Define Engagement States

Answer 37

When using Frequency as the strategy
1. Core, Your core should be defined based on your natural frequency definition
2. Power/Casual, build hypotheses for power and casual by looking at higher and lower frequencies then core.
3. Correlation Analysis, verify with a correlation analysis connecting it to long-term retention

When using Feature Usage as the strategy
1. Core, Your core should be informed by the qualitative definition of core action/problem
2. Feature Buckets, Doing a correlation analysis on the features used that lead to long-term retention.
3. # of Features, Doing a correlation analysis on the # of features used.

When using Intensity (Time, Money, Action) as the strategy
1. Core, Your core should be defined by the qualitative definition of a healthy retained user. The problem we are solving.
2. Power/Casual, Build hypotheses for power and casual by looking at higher and lower intensities.
3. Correlation, verify with a correlation analysis connecting it to long-term retention.

Question 38

Analyzing Engagement States - High-Level Analysis

Answer 38

1. Current status, What is the distribution of my engagement states?
2. How have my engagement buckets trends over time?
3. How many users are transitioning between states?

Question 39

Analyzing Engagement States
- Analyzing Engagement Further

Answer 39

1. Engagement Cohorts
2. Segmenting Engagement States

Question 40

Engagement Strategies - Identifying Engagement Opportunities

Answer 40

1. Mindset, what is the difference in mindset between engagement states? The mindsets inform the who, the message, the how, and more…
2. Pathways, what is needed for users to transition between states?
   2.1 Pathway comparison matrix. Recent transition successes and recent transition failures
3. Data Signals, how many users with this mindset?
   3.1 If you can’t represent the qualitative with a metric, try sample surveys to estimate the size of the mindset.
4. Value, what is the size of the business value for this pool of users?
   # of users in that pool * value of state transition = value of opportunity

Question 41

Engagement Strategies - Use Case And Feature Strategy

Answer 41

Considering the forgeable zone

Users can enter from many different places
Users can take many pathways from use case to use case.
Find the most common path, it can change depending on the persona
These paths normally map to the engagement states

We use feature usage when the use case is enabled by a set of different but complementary features. They are not competing with each other, for the user’s time for example.

5 Steps to transition engagement states for use cases or feature usage
1. Signal, what is the signal in the data that tells us they are a good candidate? Who, Past, what have they done? Present.
2. Real Estate, what is the potential real estate I can use to engage them? 2.1 Notifications, transactional emails, think about real estate in your transactional emails that already have the attention of the user.
2.2 Website, homepage, login page, help, blog
2.3 Environment
2.4 End States, end states and other dead spots in the user flows of your product.
2.5 Highlights
2.6 Promo Spots
3. Message, what is the message or hook that gets them to engage?
4. Activation Flow, what is the path from exposure to habit? Similar to activating a new user
5. Destination, what is the destination and definition of success?

Question 42

Engagement Strategies - Frequency Strategy

Answer 42

4 steps to increase frequency
1. Optimize Core Loop, manufactured + environment loops
2. Add Supplemental, build new supplemental loops that reinforce the core loops by identifying new manufactured or environment triggers, Finding supplemental loops comes down to establishing/finding new triggers
3. Optimize Supplemental
4. Moderate, Moderate loops over time to save users from themselves. Increasing frequency doesn’t equal to spamming

Question 43

Engagement Strategies - Intensity Strategy

Answer 43

Be careful with the inverse relationship intensity and value. E.g. Productivity products.

One way to increase intensity is by increasing the frequency.

To increase the intensity. We need to make sure there are no dead ends.

Find the end states and exit paths and place something in that path that continues the experience.

Question 44

Engagement Strategies - Building An Engagement Machine

Answer 44

1. Signal, Who they are, what they have done, what they are doing?
2. Strategy, a prioritized list of engagement paths we can send them down.
3. Path. Real estate and the path we use to try and lead them to the engagement opportunity
4. Success/Fail, were we successful in getting them to establish the habit?

It is a self-reinforcing system, iteration can help us improve.

Question 45

Resurrection

Answer 45

Dormant, Chruned and Non-Activated users are different types.

Non-Activated users most likely have an activation problem not a resurrection problem.

The longer the user stays in the dormant state, the harder to resurrect them

Users * Response Rate * % Retained = Lift

Be aware of the
1. Damage to Your Channels
2. Negative Word of Mouth

When to use resurrection?
1. After activation and Engagement, with limited resources, bigger impact can come from activation and engagement initiatives
2. Understand the true opportunity and cost, Dig to understand the true opportunity and potential cost of resurrection efforts.
3. Around naturally high psych time periods, Capitalize on naturally high psych time periods for your product and audience.

Naturally high psych moments can occur based on seasonal adoption or buying cycles. E.g. January for a fitness app.

Humans look to justify their decisions not change them.

The best resurrection strategy is the prevention

Question 46

Resurrection - Define At-Risk Users

Answer 46

1. Key Actions, identifying key actions in the product that are clear signals of at-risk users. Might be too late.
   1.1 Qualitative Responses, low Net Promotor Score, NPS
   1.2 Customer Support Activity
   1.3 Cancel Flows
   1.4 Exports data
2. Direction Of Engagement, Looking at the negative direction of engagement
3. Prediction Model, Build a model that gives each user a score of how likely they are to churn.
   3.1 Define Formant
   3.2 Hypotheses, what are possible hypotheses of things that might lead to a user going dormant?
   3.3 Correlation, what are the variables that have a positive correlation with dormant users?
   3.4 Model, what is the probability that a user will go dormant in some time period?
   3.5 Predict + Refine, how accurate are my predictions? FP, FN

Data + effort required increase from 1 to 3

If you are successful in intervening, the model will become obsolete.

Question 47

Resurrection - Define Dormant Users

Answer 47

1. Define Dormant, when do you consider someone dormant vs casual?
2. Voluntary vs Involuntary, how have uses gone dormant for voluntary or involuntary reasons?
3. Satisfied vs Unsatisfied, within the voluntary segment, have they gone dormant because they are satisfied or unsatisfied?

Question 48

Resurrection - Define Churned Users

Answer 48

1. Historical Dormant Users, how many users have ever been dormant for X periods or longer?
2. Current Dormant Users, How many users are currently dormant for X periods or longer?
3. Estimate Percentage Of No Return, What percentage are unlikely to return by dormancy age? 2/1 = the percentage
4. Qualitative Judgement, where do we want to draw the line between dormant state and churned state?

Question 49

Resurrection - Estimating The Cost

Answer 49

1. The influence of WOM for your audience + product, WOM plays a stronger role in the decision process for some audiences and products.
2. The sensitivity of the channel, Some channels are easier to damage than others.
3. Company’s ability to absorb if something goes wrong, if something goes wrong with the channel, what is the ability to absorb?

Question 50

Resurrection Strategies

Answer 50

Can help us fix the problem from ok to good retention.

Question 51

Resurrection Strategies - For At-Risk Users

Answer 51

The best resurrection strategy is the prevention

1. Habit Reinforcement, Reinforcing the “why” of the product.
2. Use Case Transition, Transitioning a user to a different use case for the product. Disengaging due to completing a use case.
   2.1 Same 5 steps to transition engagement states for use case or feature, signal -> real estate -> message -> activation flow -> destination
3. Anti-Conversion, Increasing the physical and cognitive friction to make them question leaving the product.
   3.1 Flipping Conversion Rate Optimization (CRO), From remove friction to add friction. From don’t make the user think to make the user think. From motivate towards action to motivate away from the action.
   3.2 Balancing how aggressive of a strategy

Question 52

Resurrection Strategies - Involuntary Dormant Users

Answer 52

1. Product Issues, Technical bugs, users getting logged out, switched devices, user switched emails, etc
2. Leaving, Left company, market, or made some other change where they can no longer use the product
   2.1 Collect secondary contact
3. Payment, Payment failure such as expired credit card, limit reached, address change, etc.

Question 53

Resurrection Strategies - Voluntary Dormant Users

Answer 53

1. Why, why has the person gone dormant?
   1.1 Received the value, but not a repeatable use case.
   1.2 Product didn’t do what they wanted it to do.
   1.3 Over-promised, under-delivered
   1.4 Unsustainable habit loop, habit became too much and a reason to quit
   1.5 Pricing movement
   1.6 Competitor movement
2. Message, what is the message/offer that will get them to change their mind?
   2.1 New news
   2.2 New use case
   2.3 Incentive for repeat behaviour
   2.4 +Psych Moment, positive psych moment generated by activity, time or some other event
3. Timing, what is the right time to send someone the offer?
   3.1 Activity Based, activity based on some action in the product by the user or some other user in their network/instance.
   3.2 Timed, timed sequence based on the date going dormant
   3.3 Campaign, Campaign driven by some outside event such as a holiday, new feature, etc
4. Channel, what channels are available to reach them?
5. Re-Activation, what is the quickest path to re-establishing the habit?

Question 54

ICED Theory For Infrequent Products

Answer 54

Infrequency: Infrequent <—> Frequent
Control: No Control Over Experience <—> Control Over Experience
Engagement: Low Engagement <—> High Engagement
Distinctiveness: Not Distinctive <—> Distinctive

For frequent products, the product-market fit is defined by the retention of users. And the growth is various Growth Loop concepts
For infrequent products, the product-market fit is a function of market penetration. And the growth is ICED theory in conjunction with Growth Loops

Question 55

Loyalty Within ICED Theory

Answer 55

1. Repeat Transaction, Customers keep coming back for more transactions
2. Share of Wallet, Customers spend more on average over time
3. Advocacy, Customers say good things about your product to others

Question 56

ICED Theory - Engagement Attributes

Answer 56

1. Complexity, failure to correctly manage product complexity results in disloyalty. Buying a property is a complex transaction process.
2. Degree of Touch, Products with a single touch may not provide reinforcement of product experience or a strong brand recall.
3. Predictability of Retention, If product retention is unpredictable, then managing the predictability of retention is crucial

Question 57

Market Penetration

Answer 57

A business’ market penetration measures how much their product is being sold relative to the total estimated market for that product, expressed as a percentage. Also known as market penetration rate.

Market penetration rate = (number of customers / target market size) * 100

It’s suggested that the average market penetration for a consumer product is 2 to 6%, while business products B2B can range anywhere from 10 to 40%. IF your SaaS solution captures 10% of your TAM, you will probably be doing quite well.

Question 58

Light Control To Heavily Managed Business Models

Answer 58

Light -> Heavy
1. SaaS, Turbotax
2. SaaS-Like Network, Better
3. Light Market, Zillow, Thumbtack, craigslist
4. Managed Marketplace, Etsy, airbnb
5. Heavily-Managed Marketplace, Lenskart
6. Vertically Integrated, OYO, Stitch Fix

Question 59

Retention Key Terminologies

Answer 59

Lifetime retention: This refers to the number of times a user returns during their lifetime to perform transactions. For Example: A TurboTax user returns every year to file their taxes.

Active retention: This refers to the number of times a user returns to perform a single transaction. Within a tax season, a user returns multiple times to complete a transaction.

Infrequency (to frequency) dimension of ICED Theory: When we transition from infrequency to frequency, we are enabling lifetime retention through increased frequency. For example, an application such as Zillow Zestimate facilitates lifetime retention making people come back every quarter or six months to the product.

Engagement dimension of ICED Theory: This is primarily meant to boost lifetime retention of performing the core transaction, while also having some impact on active retention.

Question 60

ICED Theory’s Hierarchy Of Engagement

Answer 60

1. Increase the No. of users with key actions
2. Captivate customers by planting loyalty hooks
3. Enhance customer loyalty

For Infrequent Products WITH Retention
1. Focus on growing No. users completing the core actions
2. Captivate customers by increasing switching cost
3. As users engage, their retention with the product improves
Manage Complexity increases from 1 -> 3

For Infrequent Products WITHOUT Retention
1. Focus on growing No. users completing the core actions
2. Make it memorable to create a strong recall
3. As users engage, they amplify the word of mouth of the product
Towards Constant Touch increases from 1 -> 3

Question 61

Product Stack

Answer 61

1. Community, Communities exist explicitly or implicitly for various products
2. Infrastructure, external producers build or create value
3. Data, every product employs data in different ways.

Question 62

Product Vision - Get Big On, Lead, Expand (GLEe)

Answer 62

To outline the GLEe model for your product and company, ask yourself three questions:

1) What is the initial product that enables the company and product to “Get Big” over the first 3–5 years of its life? Are there trends that the product can “surf,” much like Netflix rode the wave of DVD players and e-commerce? What are these trends for your company and product?

2) What will you “Lead?” Three to five years in the future, what is the next wave your product or company will ride — the equivalent of internet video for Netflix?

3) Once your product establishes a leadership position, how might it “Expand” even further? Given the brand, network effects, economies of scale, and unique technology your product will have at this point, what is the next wave of activity?

Question 63

Overview Of ICED Theory

Answer 63

1 GLEe
1.1 Ingrequency
1.1.1 Move to timing known–easy to influence
1.1.2 Add adjacent use cases
1.1.3 Latch on to/layer a habit product
1.2 Control Over Experience
1.2.1 Close the loop
1.2.2 Increase the score product’s convenience
1.2.3 Move toward complete control over experience
1.3 Engagement
1.3.1 Plant Loyalty Hook
1.3.1.1 Make it personal by make users invest
1.3.1.2 Engineer peak moments
1.3.1.3 Balancing supply and demand
1.3.2 Manage complexity
1.3.2.1 Increase trust in complex transactions
1.3.2.2 Defang difficult decision-making
1.3.2.3 Reduce cognitive load and effort load
1.3.3 Increase touch from single to constant
1.3.3.1 Expand touch points across channels
1.3.3.2 Add use case to move to constant touch
1.3.3.3 Flank the product line
1.4 Distinctive
1.4.1 Reconfigure product stack
1.4.2 Reconfigure value chain
1.4.3 Reconfigure to ensure supply-side advantage (exclusive and scarcity)

Question 64

PMD Parameters

Answer 64

1 Non-Penetrable vs Penetrable
1.1 High infrequent products are generally of high-ticket value vs Products with low infrequency are of low order value
1.2 It may not be applicable to everyone in the market vs Could be applicable to most users in the market
1.3 Buying a car or property vs Shoes headphones

2. Macro Factors Non-Resilient vs Resilient
   2.1 Some products move with ebbs and flows of an economy vs Some products have constant demand throughout everything
   2.2 It may not be applicable to everyone in the market vs it could be applicable to most people in the market
   2.3 Buying a car or property vs Health, Taxes and Education.
3. Impulsive vs Deliberate
   3.1 Products that are low-ticket value, simple to use vs Products that are of high-ticket value or complex
   3.2 It could be applicable to most people in the market vs it may not be applicable to everyone in the market
   3.3 Shoes, headphones vs Buying a car or property

Question 65

ICED For PMD Parameters

Answer 65

Infrequent
Control Over Experience
Engagement
Distinctive
Summary

For Lack of Penetrability (P)
Moving to frequency makes it penetrable as the product scope is expanded
Gain control over experience to maximize the revenue
-
-
[IC]

For Macroeconomy Resilience (M)
Making it frequent provides resilience against macro factors (Most habit-products are resilient)
-
-
Being Distinctive may help acquisition depending on the type of the product
[ID]

Provides convenience to transact with a single point of contact
Defang the decision-making by dialing up on the engagement
Being distinctive means a strong brand and creates strong pull factor while making decisions
[CED]

Question 66

ICED Theory Assessment & Strategy Evaluation

Answer 66

Rows: Current Status (L/M/H), Ability To Move (Easy/M/Hard), Check List on Which Strategy is Relevant, Impact of Strategy (L/M/H), Risk to Manage (Viability, Adoption, Execution), Move or Stay

Columns: Frequency, Control Over Experience, Managing Complexity, Extent of Touch, Strength of Hierarchy of Engagement (loyalty and trust), Distinctiveness