How to Keep Users Coming Back to Your App in 2020

Item not bookmarked
Resource bookmarked
Bookmarking...
⛏️
Guest Miner:
Sylvain Gauchet
Review star
Review star
Review star
Review star
Review star
💎  x
6

Brittinee S. Phillips (Marketing Consultant) discusses with Nicholas Beck (Director of Product at K Health), Nataliya Novikova (Director, Product Marketing & CRM at Grubhub, previously Audible) and Adithi Sampath (at Walmart Labs) about strategies and tactics to improve onboarding and app engagement.

Source:
How to Keep Users Coming Back to Your App in 2020
(no direct link to watch/listen)
(direct link to watch/listen)
Type:
Panel
Publication date:
June 17, 2020
Added to the Vault on:
June 18, 2020
Invite a guest
These insights were shared through the free Growth Gems newsletter.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
💎 #
1

Defer asking users to opt in to when the context is right, when there is a reason/value for it. Example: getting notified when your doctor is available. If it's to complicated to explain in your onboarding, defer.  

12:35
💎 #
2

Experiment with deferring some onboarding "asks", especially if you know what the intent of the user is coming in. You can even turn that into the concept of a "post board" where you ask users things after they have completed what they came in to do and throughout the experience. 

14:16
💎 #
3

Think broader than just the first transaction: think at the brand/app level. After that first activity (audiobook, delivery), the first few weeks of onboarding are crucial to create habits. 

19:34
💎 #
4

LTV can and will change so you need to look back at what the customers actually generated compared to your LTV, which is rarely done. 

39:09
💎 #
5

To identify invested/engaged users you can look at how often your app is used but also at users' "wishlists" (likes/dislikes, favorites, etc.) because it increases the "cost" of switching products. To identify a further level of engagement you can also look at the usage metrics of the features built around these "wishlists" (e.g. reordering from a restaurant). 

54:18
💎 #
6

Look for what provides the best customer experience for a power user (the experience you want to create, what is important for users) and use that to review your first-time onboarding based on that. 

58:15
The gems from this resource are only available to premium members.
💎 #
1

Defer asking users to opt in to when the context is right, when there is a reason/value for it. Example: getting notified when your doctor is available. If it's to complicated to explain in your onboarding, defer.  

12:35
💎 #
2

Experiment with deferring some onboarding "asks", especially if you know what the intent of the user is coming in. You can even turn that into the concept of a "post board" where you ask users things after they have completed what they came in to do and throughout the experience. 

14:16
💎 #
3

Think broader than just the first transaction: think at the brand/app level. After that first activity (audiobook, delivery), the first few weeks of onboarding are crucial to create habits. 

19:34
💎 #
4

LTV can and will change so you need to look back at what the customers actually generated compared to your LTV, which is rarely done. 

39:09
💎 #
5

To identify invested/engaged users you can look at how often your app is used but also at users' "wishlists" (likes/dislikes, favorites, etc.) because it increases the "cost" of switching products. To identify a further level of engagement you can also look at the usage metrics of the features built around these "wishlists" (e.g. reordering from a restaurant). 

54:18
💎 #
6

Look for what provides the best customer experience for a power user (the experience you want to create, what is important for users) and use that to review your first-time onboarding based on that. 

58:15
The gems from this resource are only available to premium members.

Gems are the key bite-size insights "mined" from a specific mobile marketing resource, like a webinar, a panel or a podcast.
They allow you to save time by grasping the most important information in a couple of minutes, and also each include the timestamp from the source.

💎 #
1

Defer asking users to opt in to when the context is right, when there is a reason/value for it. Example: getting notified when your doctor is available. If it's to complicated to explain in your onboarding, defer.  

12:35
💎 #
2

Experiment with deferring some onboarding "asks", especially if you know what the intent of the user is coming in. You can even turn that into the concept of a "post board" where you ask users things after they have completed what they came in to do and throughout the experience. 

14:16
💎 #
3

Think broader than just the first transaction: think at the brand/app level. After that first activity (audiobook, delivery), the first few weeks of onboarding are crucial to create habits. 

19:34
💎 #
4

LTV can and will change so you need to look back at what the customers actually generated compared to your LTV, which is rarely done. 

39:09
💎 #
5

To identify invested/engaged users you can look at how often your app is used but also at users' "wishlists" (likes/dislikes, favorites, etc.) because it increases the "cost" of switching products. To identify a further level of engagement you can also look at the usage metrics of the features built around these "wishlists" (e.g. reordering from a restaurant). 

54:18
💎 #
6

Look for what provides the best customer experience for a power user (the experience you want to create, what is important for users) and use that to review your first-time onboarding based on that. 

58:15

Notes for this resource are currently being transferred and will be available soon.

Onboarding, intent and asking for permissions

Nicholas (K Health)

[💎@12:35] Defer asking users to opt in to when the context is right, when there is a reason/value for it.  Example: getting notified when your doctor is available. If it's to complicated to explain in your onboarding, defer.


[💎@14:16] Experiment with deferring some onboarding "asks", especially if you know what the intent of the user is coming in. You can even turn that into the concept of a "post board" where you ask users things after they have completed what they came in to do and throughout the experience.


Adithi Sampath (Walmart Labs)

Do timely prompts for education and awareness. Ask for basic information upfront and anything else, keep that for later.

Focus on having an intuitive interface as well. Example for e-commerce: save sign in for right before checkout instead of when they come in, or even having a guest checkout (and ask for registration when they come back).

On the other hand asking for location can be crucial to check item availability in a local store for example. But still tell them why you're asking this.


Nataliya Novikova (Grubhub)

Keep in mind what user has in mind when coming to your app. Example: a specific audiobook or wanting to browse.

In general, show value first.

[💎@19:34] Think broader than just the first transaction: think at the brand/app level. After that first activity (audiobook, delivery), the first few weeks of onboarding are crucial to create habits.

Personalization based on collected data

Adithi Sampath (Walmart Labs)

If it's a new user you might not have the right information (yet). So you can look at "similar" users (e.g. same area) to find popular items.

This can start a virtuous cycle of having users purchase so you do end with data about the user.

You need to really understand the cohorts and the segments of users.

Cohort profiling and personalization with privacy and GDPR?

  • Nicholas - always think about which channels to use and communicate (and what they're allowed to use). They have a free version where you can answer questions about your condition but it is done in an anonymous fashion and used by their algorithm to provide information.
  • When using push/SMS they need to make sure they do not expose user's personal data.

Calculating LTV

  • Nataliya - cares a lot about incrementality. A lot of decisions are based on LTV (like how much you want to spend to acquire customers) and how to drive it up.
  • [💎@39:09] LTV can and will change so you need to look back at what the customers actually generated compared to your LTV, which is rarely done.
  • Determined based on historical data and complex model that the data science team used.
  • Adithi - LTV is a medium/long term metric. The key aspect of figuring out LTV is to understand what your retention rate looks like for different cohorts of customers and requires a model. The other parts like average order value and frequency are more straightforward.
  • Nicholas - At previous meal subscription company LTV was a function of number of orders/deliveries, number of meals as well as the cost to acquire them (?).

When do you need to start using AI (and not just BI)

  • Adithi - 1. When you can not reduce things to a rule-based model 2. if you want to scale fast 3. If you need things to happen in real time.
  • Nicholas - you need to make sure that you have the team in place to adapt the model (including with human input)

User metrics to identify interested vs. invested users

  • Nicholas - different levels of engagement 1. Anonymous users that go to the app and put in their symptoms, which can happen several times 2. Registered free users that decide to create an account (asked after a few "cases" are built) 3. Registered paid users, with different levels of memberships depending on how engaged you are.
  • Being willing to give personal medical information, regardless if you're a free or paid user, also shows engagement.
  • Adithi - an interested user is someone that lands on the app/site. At the session level, an invested user is someone that adds something to basket and showing intent to check out. At a broader level it's about retention and how much people come back.
  • Nataliya - [💎@54:18] To identify invested/engaged users you can look at how often your app is used but also at users' "wishlists" (likes/dislikes, favorites, etc.) because it increases the "cost" of switching products. To identify a further level of engagement you can also look at the usage metrics of the features built around these "wishlists" (e.g. reordering from a restaurant).

One last tip

  • Nicholas - Push notifications are really powerful. To help users accept, make sure you use a custom prompt explaining why and the value it will provide before the prompt.
  • Adithi - Take the time to make the difference what are vanity metrics vs. what are important metrics.
  • Nataliya - [💎@58:15] Look for what provides the best customer experience for a power user (the experience you want to create, what is important for users) and use that to review your first-time onboarding based on that.



The notes from this resource are only available to premium members.

Onboarding, intent and asking for permissions

Nicholas (K Health)

[💎@12:35] Defer asking users to opt in to when the context is right, when there is a reason/value for it.  Example: getting notified when your doctor is available. If it's to complicated to explain in your onboarding, defer.


[💎@14:16] Experiment with deferring some onboarding "asks", especially if you know what the intent of the user is coming in. You can even turn that into the concept of a "post board" where you ask users things after they have completed what they came in to do and throughout the experience.


Adithi Sampath (Walmart Labs)

Do timely prompts for education and awareness. Ask for basic information upfront and anything else, keep that for later.

Focus on having an intuitive interface as well. Example for e-commerce: save sign in for right before checkout instead of when they come in, or even having a guest checkout (and ask for registration when they come back).

On the other hand asking for location can be crucial to check item availability in a local store for example. But still tell them why you're asking this.


Nataliya Novikova (Grubhub)

Keep in mind what user has in mind when coming to your app. Example: a specific audiobook or wanting to browse.

In general, show value first.

[💎@19:34] Think broader than just the first transaction: think at the brand/app level. After that first activity (audiobook, delivery), the first few weeks of onboarding are crucial to create habits.

Personalization based on collected data

Adithi Sampath (Walmart Labs)

If it's a new user you might not have the right information (yet). So you can look at "similar" users (e.g. same area) to find popular items.

This can start a virtuous cycle of having users purchase so you do end with data about the user.

You need to really understand the cohorts and the segments of users.

Cohort profiling and personalization with privacy and GDPR?

  • Nicholas - always think about which channels to use and communicate (and what they're allowed to use). They have a free version where you can answer questions about your condition but it is done in an anonymous fashion and used by their algorithm to provide information.
  • When using push/SMS they need to make sure they do not expose user's personal data.

Calculating LTV

  • Nataliya - cares a lot about incrementality. A lot of decisions are based on LTV (like how much you want to spend to acquire customers) and how to drive it up.
  • [💎@39:09] LTV can and will change so you need to look back at what the customers actually generated compared to your LTV, which is rarely done.
  • Determined based on historical data and complex model that the data science team used.
  • Adithi - LTV is a medium/long term metric. The key aspect of figuring out LTV is to understand what your retention rate looks like for different cohorts of customers and requires a model. The other parts like average order value and frequency are more straightforward.
  • Nicholas - At previous meal subscription company LTV was a function of number of orders/deliveries, number of meals as well as the cost to acquire them (?).

When do you need to start using AI (and not just BI)

  • Adithi - 1. When you can not reduce things to a rule-based model 2. if you want to scale fast 3. If you need things to happen in real time.
  • Nicholas - you need to make sure that you have the team in place to adapt the model (including with human input)

User metrics to identify interested vs. invested users

  • Nicholas - different levels of engagement 1. Anonymous users that go to the app and put in their symptoms, which can happen several times 2. Registered free users that decide to create an account (asked after a few "cases" are built) 3. Registered paid users, with different levels of memberships depending on how engaged you are.
  • Being willing to give personal medical information, regardless if you're a free or paid user, also shows engagement.
  • Adithi - an interested user is someone that lands on the app/site. At the session level, an invested user is someone that adds something to basket and showing intent to check out. At a broader level it's about retention and how much people come back.
  • Nataliya - [💎@54:18] To identify invested/engaged users you can look at how often your app is used but also at users' "wishlists" (likes/dislikes, favorites, etc.) because it increases the "cost" of switching products. To identify a further level of engagement you can also look at the usage metrics of the features built around these "wishlists" (e.g. reordering from a restaurant).

One last tip

  • Nicholas - Push notifications are really powerful. To help users accept, make sure you use a custom prompt explaining why and the value it will provide before the prompt.
  • Adithi - Take the time to make the difference what are vanity metrics vs. what are important metrics.
  • Nataliya - [💎@58:15] Look for what provides the best customer experience for a power user (the experience you want to create, what is important for users) and use that to review your first-time onboarding based on that.



The notes from this resource are only available to premium members.

Onboarding, intent and asking for permissions

Nicholas (K Health)

[💎@12:35] Defer asking users to opt in to when the context is right, when there is a reason/value for it.  Example: getting notified when your doctor is available. If it's to complicated to explain in your onboarding, defer.


[💎@14:16] Experiment with deferring some onboarding "asks", especially if you know what the intent of the user is coming in. You can even turn that into the concept of a "post board" where you ask users things after they have completed what they came in to do and throughout the experience.


Adithi Sampath (Walmart Labs)

Do timely prompts for education and awareness. Ask for basic information upfront and anything else, keep that for later.

Focus on having an intuitive interface as well. Example for e-commerce: save sign in for right before checkout instead of when they come in, or even having a guest checkout (and ask for registration when they come back).

On the other hand asking for location can be crucial to check item availability in a local store for example. But still tell them why you're asking this.


Nataliya Novikova (Grubhub)

Keep in mind what user has in mind when coming to your app. Example: a specific audiobook or wanting to browse.

In general, show value first.

[💎@19:34] Think broader than just the first transaction: think at the brand/app level. After that first activity (audiobook, delivery), the first few weeks of onboarding are crucial to create habits.

Personalization based on collected data

Adithi Sampath (Walmart Labs)

If it's a new user you might not have the right information (yet). So you can look at "similar" users (e.g. same area) to find popular items.

This can start a virtuous cycle of having users purchase so you do end with data about the user.

You need to really understand the cohorts and the segments of users.

Cohort profiling and personalization with privacy and GDPR?

  • Nicholas - always think about which channels to use and communicate (and what they're allowed to use). They have a free version where you can answer questions about your condition but it is done in an anonymous fashion and used by their algorithm to provide information.
  • When using push/SMS they need to make sure they do not expose user's personal data.

Calculating LTV

  • Nataliya - cares a lot about incrementality. A lot of decisions are based on LTV (like how much you want to spend to acquire customers) and how to drive it up.
  • [💎@39:09] LTV can and will change so you need to look back at what the customers actually generated compared to your LTV, which is rarely done.
  • Determined based on historical data and complex model that the data science team used.
  • Adithi - LTV is a medium/long term metric. The key aspect of figuring out LTV is to understand what your retention rate looks like for different cohorts of customers and requires a model. The other parts like average order value and frequency are more straightforward.
  • Nicholas - At previous meal subscription company LTV was a function of number of orders/deliveries, number of meals as well as the cost to acquire them (?).

When do you need to start using AI (and not just BI)

  • Adithi - 1. When you can not reduce things to a rule-based model 2. if you want to scale fast 3. If you need things to happen in real time.
  • Nicholas - you need to make sure that you have the team in place to adapt the model (including with human input)

User metrics to identify interested vs. invested users

  • Nicholas - different levels of engagement 1. Anonymous users that go to the app and put in their symptoms, which can happen several times 2. Registered free users that decide to create an account (asked after a few "cases" are built) 3. Registered paid users, with different levels of memberships depending on how engaged you are.
  • Being willing to give personal medical information, regardless if you're a free or paid user, also shows engagement.
  • Adithi - an interested user is someone that lands on the app/site. At the session level, an invested user is someone that adds something to basket and showing intent to check out. At a broader level it's about retention and how much people come back.
  • Nataliya - [💎@54:18] To identify invested/engaged users you can look at how often your app is used but also at users' "wishlists" (likes/dislikes, favorites, etc.) because it increases the "cost" of switching products. To identify a further level of engagement you can also look at the usage metrics of the features built around these "wishlists" (e.g. reordering from a restaurant).

One last tip

  • Nicholas - Push notifications are really powerful. To help users accept, make sure you use a custom prompt explaining why and the value it will provide before the prompt.
  • Adithi - Take the time to make the difference what are vanity metrics vs. what are important metrics.
  • Nataliya - [💎@58:15] Look for what provides the best customer experience for a power user (the experience you want to create, what is important for users) and use that to review your first-time onboarding based on that.