How can app store conversion rates be optimized and what impact can those rates have on overall CPI?

Question

What is a good store conversion rate and how much can it be optimized? Is it more or less dependent on the game itself or can it be raised a significant amount through ASO testing?

I know google has the option to do store optimization tests with icons and screenshots. Is it best practice to do store tests on google play and apply them to iOS or better to use an ASO specific tool like SplitMetrics?

Does store conversion rate play a huge role in overall CPI for a game or is it still mostly down to performance of creatives and game itself?

Answer 1

• What is a good store conversion rate?

In my opinion, one that is higher than you had before previously: keep testing.
Google provides category benchmarks by country within the Console. I'd start prioritizing creative testing if you are below the 50th percentile.. once you get over 75th, gains are likely to be smaller to non-existent. Apple doesn't provide such benchmarks, but AppFollow and AppTweak both do on their websites. Make sure you compare similar category, country and very importantly same segment (search vs browse vs app referrer...)

 

• how much can it be optimized? Is it more or less dependent on the game itself or can it be raised a significant amount through ASO testing?

Depends where you're starting from really, but if you're below benchmark or haven't iterated much, there are likely huge gains to get. Small gains in conversion rate can mean big wins for both organic and even more importantly for paid UA, as higher CR improves your position in the auction...

• I know google has the option to do store optimization tests with icons and screenshots. Is it best practice to do store tests on google play and apply them to iOS or better to use an ASO specific tool like SplitMetrics?

As it's free & simple, definitely worth a shot trying Google Play Experiment. That said, I wouldn't trust that data blindly, certainly not without many thousands installs per variant in all cases (confidence intervals are very poor and the underlying data isn't solid either, plus who knows how good of a traffic mix Google does, disclosed metrics are vastly incomplete).
Great source of input, but poor decision-making source.

In any case, I wouldn never port those Play results onto the AppStore: the layout, UX and user profiles differ massively. TheTool has a recent article on their blog on why not conclude for iOS what seems to work on Android.

Using a "fake appstore" is definitely a good complement to the Play Store Experiment. It's not perfect/ideal either, but gives you a secod input, with extremely valuable details on user behaviour on the page that no other methods can provide. Providers include Splitmetrics, Storemaven & ASOgiraffe, all 3 with very good content on their blog/site.

There are at least 2 other testing methods: sequential (measuring before vs after the change), and Apple Search Ads creative sets (now available on brand campaigns).  All 4 methods have their own flaws, I'd recommend comparing insights from at least 2 before concluding anything, unless you're early stage and need velocity, or other departments (brand, HIPPO etc) are pushing for something fast & dirty.

Beyond all sources listed above, I'd highly recommend this very complete post from ASOstack which details those method but also explain their detailed process on how to run those creative tests..
Luca Giacomel's open source Sherlock can help you automate AppStore Connect data analysis when updating creatives.
You can also check these case study:
- SmartNews / Google Play Experiment
- Wallapop / StoreMaven
 

• Does store conversion rate play a huge role in overall CPI for a game or is it still mostly down to performance of creatives and game itself?

Install conversion rate plays a huge role in UA overall and CPI in particular. Most networks will convert your CPI/CPA/ROAS targets into CPM, and the install rate is one of the component of the equation. Only very unsophisticated networks would look at CTR alone... Do iterate on your creatives, do improve your product conversion rate, but don't let the install rate without iterations if you're serious about acquiring users. Also keep a look on the whole top funnel together (CTR * IR) not independently.
 

Answer 2

What is a good store conversion rate and how much can it be optimized? Is it more or less dependent on the game itself or can it be raised a significant amount through ASO testing?

This will vary immensely from app to app and between different categories, and how much traffic your app is getting, and from where.

Apps that get nothing but search traffic installs will have much higher install rates than apps that have a high-volume user acquisition cannon shooting a torrent of cheap eyeballs at their store page.

I personally count 30% as decent, 50% as great, and 70%+ as amazing on the games we make, but certain traffic sources can crush those numbers down into the 5-25% range which can look really scary.

The main impactor your install rate is going to be the actual appearance of your app store landing page, and changing assets can change your install rate dramatically. When I run ASO experiments with new assets, I'm typically looking for a change of at least 5% more conversions than before.
 

I know google has the option to do store optimization tests with icons and screenshots. Is it best practice to do store tests on google play and apply them to iOS or better to use an ASO specific tool like SplitMetrics?

If you're just starting out, I'd say that testing on Android and applying on iOS is totally valid. Some people will tell you that it doesn't work, but those people tend to work for SplitMetrics.

The great benefit of using Google's integrated A/B testing is that it applies your variants across the entire store experience so you get to test with actual organic traffic. This means that you don't have to buy extra traffic to go to your store page.

Third party tools like StoreMaven and SplitMetrics are way too expensive for what they offer and if you use them you need to buy traffic to the test - which means you can only test the conversion rate of your paid users.

 

Does store conversion rate play a huge role in overall CPI for a game or is it still mostly down to performance of creatives and game itself?

Store conversion plays an enormous role in CPI! If you're paying $10 to reach 1000 people with your ads, get a 1% click through (10 people), and a 10% install rate, you've got 1 install at $10 CPI. If you just raise your install rate from 10% to 30% that would reduce your CPI from $10 to $3.33, which might make the difference between your business working or going bust.