Scott Ritchie
Head of Business Development at Pyze
Where to hear and read about Scott Ritchie1 episode or test
Scott Ritchie is Head of Business Development at Pyze, a mobile analytics company focused on app engagement and retention. The thing most worth learning from him is that retention sets how much of your addressable market you keep, so build behavior tracking, segmentation and automated engagement before buying more users.
You're not really eliminating the data scientists, but what you're doing is you're applying them to understand the business rules much more than the technical rules.
Mobile App Engagement and Retention
Scott Ritchie: the story and the lessons
1Why Scott Ritchie says retention sets your market size
Ritchie models retention against the addressable market before buying users. His model assumes a 100 million addressable market, a 10% conversion rate and about five shares per user. At 90% retention the app reaches capacity, and at 80% the market settles near 43 million users.
With no new acquisition spend, users decay. A weaker case leaves about 17 million, and roughly 36 months after the peak the app is close to empty.
He points out that acquisition gets about 90% of the attention. Ritchie says most conference talk and budgets go to acquisition, while the users already paid for leak out. He advises shifting effort to keeping them.
His words: "that's what ninety percent talk about an acquisition".
Ritchie splits growth into four stages, and says most teams work on one. The stages are acquisition, onboarding and activation (first purchase), engagement, and retention with loyalty and referrals. Analytics drives each stage.
He says machine learning can cover engagement and retention.
Ritchie warns against launching new apps to replace lost users. Many casual game and utility makers lose users from the core experience, then build more apps to refill the funnel. He says to fix retention in the core app instead.
Casino apps are the exception, because real money keeps players and retention is already huge.
2How Ritchie tracks behavior and personalizes
Ritchie tracks user actions, not features. He instruments clicks and decision points inside screens and content. New features should match how the current base already behaves, starting from the metric a feature should move.
His phrasing: "track user actions right and not features". He says focusing only on features loses your existing users.
He wants detailed attributes on events rather than many bare events. Teams often instrument too many events without defining attributes for each. Detailed attributes show how users engage with each event and steer product direction in real time.
A gaming client defined very detailed attributes and used them to drive product work for its current users. A new event can later be attributed to a segment or cohort.
Ritchie ties behavior to a user ID and carries traits across devices. Traits should follow a login across web, mobile and mobile web, with CRM data added where available. Combining iOS and Android activity makes segments reflect the whole person.
Linking usage to an exact person lets a model keep updating that person as behavior changes. He stresses that "traits are very important".
He gives each segment its own level-up path and offers. Segments progress at different speeds, so slower ones get more discounts to push them up a level. Ritchie expects that within three or four years every user will see a different app.
His example is the CNN app, where each person would open it and see different content. Games will change per player too.
3Ritchie on automating engagement with machine learning
Ritchie predicts churn and fires campaigns automatically. A model flags a user likely to leave in about three weeks to a month, such as someone who bought twice and is expected to lapse. A coupon or game offer then goes out before that happens.
He says a one-line capture means no engineers are needed to instrument.
He works from a customer value growth checklist. The order is instrumentation, usage analytics, attribution, segmentation, personalization of content and gameplay, engagement triggers, and retargeting triggers and campaigns.
Start with the metrics you want to move, then build in that order.
Ritchie keeps data scientists, but points them at business rules. Machine learning does not remove the data scientist. Their time goes to how segments level up and what offers they need, more than to dataset and modeling details.
He argues analytics can carry retention, and flags it as his belief. Ritchie says analytics with machine learning can cover engagement and retention without leaning on product development and creativity. He calls this his belief, not a proven rule.
He ties better engagement to in-app purchase revenue, such as coins, and to the value of an exit when selling the company.
Pyze's co-founder pushed an app to number one with purchased behavior data. The co-founder built an app that reached number one in its vertical. Its data came from SDKs worldwide: manufacturer, usage and which screens people view.
Ritchie notes that behavioral data like this can be bought and used to drive acquisition.
Mobile App Engagement and Retention
Acquisition gets most of the attention, but retention decides how much of your addressable market you actually keep, so build analytics, segmentation and automated engagement around real user behavior. The talk sizes the effect with retention models and gives a checklist for personalization and churn prevention.
Ritchie models retention against the addressable market before buying users. His model assumes a 100 million addressable market, a 10% conversion rate and about five shares per user. At 90% retention the app reaches capacity, and at 80% the market settles near 43 million users.
With no new acquisition spend, users decay. A weaker case leaves about 17 million, and roughly 36 months after the peak the app is close to empty.
He points out that acquisition gets about 90% of the attention. Ritchie says most conference talk and budgets go to acquisition, while the users already paid for leak out. He advises shifting effort to keeping them.
His words: "that's what ninety percent talk about an acquisition".
Ritchie splits growth into four stages, and says most teams work on one. The stages are acquisition, onboarding and activation (first purchase), engagement, and retention with loyalty and referrals. Analytics drives each stage.
He says machine learning can cover engagement and retention.
Ritchie warns against launching new apps to replace lost users. Many casual game and utility makers lose users from the core experience, then build more apps to refill the funnel. He says to fix retention in the core app instead.
Casino apps are the exception, because real money keeps players and retention is already huge.
Ritchie tracks user actions, not features. He instruments clicks and decision points inside screens and content. New features should match how the current base already behaves, starting from the metric a feature should move.
His phrasing: "track user actions right and not features". He says focusing only on features loses your existing users.
He wants detailed attributes on events rather than many bare events. Teams often instrument too many events without defining attributes for each. Detailed attributes show how users engage with each event and steer product direction in real time.
A gaming client defined very detailed attributes and used them to drive product work for its current users. A new event can later be attributed to a segment or cohort.
Ritchie ties behavior to a user ID and carries traits across devices. Traits should follow a login across web, mobile and mobile web, with CRM data added where available. Combining iOS and Android activity makes segments reflect the whole person.
Linking usage to an exact person lets a model keep updating that person as behavior changes. He stresses that "traits are very important".
He gives each segment its own level-up path and offers. Segments progress at different speeds, so slower ones get more discounts to push them up a level. Ritchie expects that within three or four years every user will see a different app.
His example is the CNN app, where each person would open it and see different content. Games will change per player too.
Ritchie predicts churn and fires campaigns automatically. A model flags a user likely to leave in about three weeks to a month, such as someone who bought twice and is expected to lapse. A coupon or game offer then goes out before that happens.
He says a one-line capture means no engineers are needed to instrument.
He works from a customer value growth checklist. The order is instrumentation, usage analytics, attribution, segmentation, personalization of content and gameplay, engagement triggers, and retargeting triggers and campaigns.
Start with the metrics you want to move, then build in that order.
Ritchie keeps data scientists, but points them at business rules. Machine learning does not remove the data scientist. Their time goes to how segments level up and what offers they need, more than to dataset and modeling details.
He argues analytics can carry retention, and flags it as his belief. Ritchie says analytics with machine learning can cover engagement and retention without leaning on product development and creativity. He calls this his belief, not a proven rule.
He ties better engagement to in-app purchase revenue, such as coins, and to the value of an exit when selling the company.
Pyze's co-founder pushed an app to number one with purchased behavior data. The co-founder built an app that reached number one in its vertical. Its data came from SDKs worldwide: manufacturer, usage and which screens people view.
Ritchie notes that behavioral data like this can be bought and used to drive acquisition.