Subscription App Stats, Push Notifications, and Fighting Churn

Notes from App Growth Summit Austin

Jacob Rushfinn

Nov 20, 2024

Hey there, it’s Jacob at Retention.Blog 👋

I got tired of reading high-level strategy articles, so I started writing actionable advice I would want to read.

Every week I share practical learnings you can apply to your business.

Last week I went to AGS Austin

Shout out to Louis Tanguay and the rest of the AGS team for putting on a stellar event!

One of the coolest parts of the event was the outdoor location. BBQ, tacos, and a campfire after-party with s’mores is hard to beat.

I guess there were a few presentations too… 😂

Below are my notes from 3 presentations I found valuable.

5 Million Subscribers Later: How to Get the Most Out of Paid Subscriptions

8 Reasons Why Push Notifications Aren't Effective

Turning Churn Into Loyalty: Data-Driven Strategies to Retain and Engage Users

Let’s start with Patrick’s presentation on, “How to Get the Most Out of Paid Subscriptions”

He was a wealth of knowledge. And I loved his focus on providing very tactical and valuable insights.

Here are some quick insights on why subscription apps are a great business:

He found that subscriptions on average generate 70%+ higher ARPU than ads for your app
And it doesn’t take that much to have a valuable product: 20k paid subscribers generates about $1m in profit
Subscription apps are a very sustainable business due to renewals. An app that is 3-5 years old generates 70% of revenue from renewals
Your incentives are aligned with your users. It’s built on a model of continuous delivery and your users are paying for the improvement of the product with their subscription

🔥Hot take alert from Patrick: Weekly subscriptions are not subscriptions since the average length is 3.5 weeks. This should be thought of more as a one-time purchase, not someone that is going to stick with you for the long term.

I like this take. I dug into weekly subscriptions and when they might make sense in this post - “Weekly Subscriptions: Should you care?”

Always have a paywall during onboarding (80% of purchases happen on day 0)

And paywall testing matters!

He sees a 2x relative growth rate of apps with over 10 paywall tests per year versus those who don’t

Don’t be afraid to put new people from different teams on experiments

It’s easy for people to get stuck in their ways and miss good ideas. By shifting who is working on what, you can get fresh ideas that aren’t biased by all the previous work

Usually, a free trial is the way to go:

He sees 15% higher LTV when using a free trial
70% of paywalls offer 2 or 3 subscription durations (monthly + annual is usually the winner)
But don’t be afraid to retest these assumptions. Patrick mentioned that for a while a monthly plan with no trial, and an annual plan with a trial was a winning test. Then 6 months to a year later he retested this and saw that a trial on both monthly and annual performed better.

Don’t be too concerned about consumers comparing prices to your competitors

Based on his user research, less than 10% of consumers price shop across apps
This makes sense because going all the way through the onboarding funnel for multiple apps to get to the paywall takes a lot of time

Retention improvements are hard but important

Retention improvements compound over time, it’s easier to focus on upper funnel metrics, but retention is critical
A 1% improvement in retention can be a 5% win on LTV

Making it easier to cancel is positive in the long run

A seamless cancellation flow preserves the relationship, increasing the chance of users returning to your app.
Consumers naturally come and go. Don’t leave a bad taste in their mouth if they want to take a break

Hard or soft paywall?

Consider a hard paywall if free users don’t create value
If free users do add value to your product through WOM, community, or some network effects, an active free user base is valuable.

Félix Boudreau of Pok Pok had a great presentation on hard paywalls at MAU this year. Check it out here.

Be wary of local maxima when running a/b tests.

It’s easy to get heads down when running tests and not look at the bigger picture.
Make sure things actually make sense in terms of the value of the product and how you’re communicating those benefits to a user.
Take a step back. Sometimes incremental changes aren’t enough, and you have to reinvent things a bit to continue making progress

Be very careful when calling winners of retention tests

You often don’t have enough information since you need to call tests in a timely manner, but retention changes sometimes only materialize over months.
Try tracking cohorts from a test over a longer time period. For example, roll out the winning variant, but track the users who were exposed to both variants to see if the results are different with 3 months (or more) of data.
You can also project what different results would mean. E.g. What would a 5% lift mean? What would an 8% lift mean? What would be successful? If that retention lift disappeared what would that mean?

When should you focus on CRM?

Follow the 80/20 rule and focus on core campaigns
The value of CRM increases with the size of your user base. The work to create one email or one push notification is always the same, but the impact is very different if you’re sending to 10k users vs 100k users.
Discounts and offers can often drive successful campaigns
So what is that 20% of campaigns that drive 80% of the impact? Paywall abandoned flows, trial cancellation flows, and subscriber cancellation/winback flows will do a lot for you

There are so many benchmarks out there, but most of the benchmark data is lower than what you should actually be targeting.

A lot of data says free trial start rates can be 3-7%, but really 15%+ is what you should aim for
For your total install-to-paying user ratio aim for 6-10%
Renewal benchmarks:
- 60-70% monthly for month 1
- 40-50% renewal for annual for year 1

Thank you Patrick for the valuable insights!

Go check him out on LinkedIn or at his company, The App Shop

Next up is Harsh with, “8 Reasons Why Push Notifications Aren't Effective”

Users don’t opt in due to a lack of trust
- 52% of users find push to be annoying

Check out my post on different push permission prompt styles here.

Push is spammy
- if you send too often users experience fatigue, but too infrequently causes no impact.
- Use behavioral analytics to target the right customers at the right time
Push notifications can easily be blocked. What are the top reasons?
- Content that’s not relevant
- Lack of personalization
- Poor timing
- Interruptive to their day
No clear CTA
- Give users a purpose that actually provides value to them.
- Great messages have a tangible benefit, easy CTA, a sense of urgency, clear and concise writing
Lack of mechanism for feedback
- If users can’t configure the type of messages they receive they might turn them all off or uninstall your app.
- A preference center will help users select the type of engagement they want
Network issues
- Have a retry mechanism in place, if it’s not delivered try a backup channel like email or SMS
- This is a little different if you’re building the solution yourself or using a 3rd party tool
Improper implementation of push
- There can be many technical issues that cause notifications not to be delivered so make sure you’re QA-ing your setup thoroughly

Last note here: A good onboarding flow that successfully sells the product will boost the effectiveness of your opt-in prompt because you’ve built trust in advance.

It seems like I didn’t get all 8 reasons…oh well.

Check out Harsh here on LinkedIn

Last, but not least, Anastasia with, “Data-Driven Strategies to Retain and Engage Users”

Understanding churn reasons is a powerful lever to loyalty. Many possible reasons for churn like:
- Performance issues
- Unmet expectations and you don’t deliver what you promised
- Poor onboarding
- Limited value over time with no recurring value
Do you know your crash rate?
- Less than 1% is acceptable, but the lower the better.
- Monitor subscription or trial cancellation rates by app version to see if you have bugs connected to certain versions
Data:
- Funnel analysis is an easy way to identify significant drop-off
- Add tracking events to all onboarding screens to understand user behavior
Onboarding: Too long? Too confusing? Is there a bug?
- Do you have a dead-end user journey?
- Give value early, don’t wait until the end of onboarding
Dogfooding: use your product!!!
- You’ll find many issues and understand your users better

All the companies I’ve worked at with great products have a culture of Dogfooding. Some places gave free credit to use the product (food ordering), others created leaderboards to show which employees used the app more. And it’s not enough to use it, make sure employees feel empowered to share their feedback and suggestions.

Screen record sessions:
- Software tools can help you do this easily
- This is usually easier on web, but there are more and more tools that can do this well on mobile apps
Talk to your users:
- Interviews and surveys to find churn reasons
- Right before you see drop off in your user funnel, add a “pulse” survey in-app to get feedback before they leave
How do I know if I have a product-market fit issue? Here are signals:
- Low user engagement
- Frequent feature requests different from the core offering
- High churn rates
Is the revenue model causing churn? Here are the options to try:
- Offer flexible payment options
- Justify subscription cost with recurring value
- Align pricing with competitor standards
- Let users experience value before you ask for too much (and ask to pay)
- Get user feedback and ask why they’re leaving!
Do you let customers cancel without leaving the app?
- Stop sending users away to cancel. This is an opportunity.
- They can re-engage
- you can ask why they’re leaving
- Give a win-back offer
- Subscription pause,
- Lower cost option,
- Change messaging based on churned user feedback and optimize the offering

Connect with Anastasia here on LinkedIn

Thanks again to everyone at AGS. Go check out all their other events around the world: appgrowthsummit.com

🆕 I’m trying a new section out:
My favorite LinkedIn posts and comments from the week

📣 Want to help support and spread the word?

Go to my LinkedIn here and like, comment, or share my posts.

Share this newsletter: