Demystifying ASO on Google Play - Real Case Examples

Keyword density has an impact on rankings

• Priceline updated their description
• [💎@06:09] On the Google Play Store, AppTweak recommends a keyword density between 1% and 2%. 3% keyword density is already high and anything above that can be considered keyword stuffing by Google.

• Comparison in keyword density before/after (lower after) thanks to the free keyword density calculator (they also have a nice Chrome extension)

• [💎@07:30] You want to look deeper into top 10 ranking progression and not just the visibility because If the app starts ranking better on high-volume keywords, it has a huge impact on visibility. If it starts ranking better on low volume keywords, the impact is not as strong.

• Teen Patti Gold (card game app) did a similar thing, this time increase keyword density for "rummy" from 0.5% to 2.1% with great visibility results on a set of 128 keywords. Focusing on the top 4 keywords with the highest volume, here are the results:

• AppTweak also helps you estimate how this translates into downloads

[💎@14:06] Because you can't have density on an infinite number of keywords you want to focus on the keywords that are the best for your app:

• Select the most relevant
• Avoid keywords that are too competitive

If you don't see an impact, you need to iterate until you do (selecting new set of keywords).

Practical tips to increase conversion

Screenshots

• The Indian app WedMeGood lost rankings vs. a new competitor "MyWed" for the "wedding planner" keyword. AppTweak started by looking at keyword density and realized WedMeGood was already doing better there. So they looked into conversion instead, starting with screenshots.
• 
  

• What MyWed's app shows in the UI for each screenshot seems more compelling: 1. Countdown 2. Checklist 3. Budget monitoring 4. Guest list manager
• Booking.com app also tested a different order in screenshots and changing the order of features changed the app conversion

Emojis can make a difference

• MyWed app had a ❤️ emoji in their short description that might have helped
• Tasty town has food emojis in both the app name and short description, which they heavily tested.
• [💎@22:17] Emojis can make a difference in conversion and you can incorporate them in your app title and/or your short description. Because you can not A/B test your app title you want to do sequential tests where you go back several times to the original emoji-less title to compare results.

• They even tested the placement of the emojis (beginning, end, middle)

Don't take video for granted

• Expedia A/B test with and without video and kept the version without video
• Sixt did a similar test and also kept the version without video
• Same with Headway

Something that is lacking here in the analysis is that adding a video on the Google Play Store replaces your first screenshots with the feature graphic as video thumbnail. So the feature graphic/thumbnail is at least as important as the video itself.

Videos don't necessarily increase conversion.

[💎@25:52] Out of 10 US apps that A/B tested video in 2020, 5/10 chose to keep the version with the video.

Q&A:

• What goes into keyword density?
• You take the keyword, check how many times it is repeated in the text vs. all keywords. The longer the text the more you need to repeat a keyword.
• Some video apps have >13% keyword density without being penalized it seems.
• Google really doesn't recommend keyword stuffing so ML really doesn't recommend it either. They've seen apps penalized with keyword density > to 4% or 5%.
• Working out the visibility of a keyword?
• Everyday look at how your app ranks everyday on these keywords then look at the next day and what happens between the two.
• How your app converts on a keyword impacts your ranking.
• How to determine which keywords to focus on?
• 1. Make sure you choose relevant keywords 2. Look at the volume and if they drive traffic 3. Look at difficulty and competition (don't go after keywords that are too competitive, use long-tail at the beginning)
• How much tinkering between updating?
• Wait at least 1 week because all apps have at least some seasonality by day. For experiments you at least want to wait until Google tells you you have significant results but still wait 7 days.
• Do you recommend using the 4,000 characters?
• [💎@37:48] Base your decision on keyword density but regarding length you need to keep the user in mind. So write what you want to say, then tweak it to have the right keyword density on important keywords.

Retention is the most determinant ranking factor

• A popular sport app in India (remains anonymous) work a lot on keyword density and conversion optimization. They noticed a huge spike in keyword rankings and in category ranking without any ASO changes. The only thing they had changed was fixing a bug that forced to uninstall/reinstall.

[💎@44:22] Retention has a huge impact on ranking in Google. Overall retention is taken into account for both keyword and category rank and Google also measures the retention at a keyword level to influence rankings.

[💎@45:04] When you select your keywords look in your search terms in the Google Play developer console and make sure to keep those with a high retention.

Impact of tags on similar apps

• The game Questland changed one of the tags for their app to see the impact it had. The change of 1 tag changed the similar apps displayed.

• It even changed on the similar apps of other apps that Questland was displayed in. It is therefore important to monitor the impact of changing google tags.

[💎@49:24] Google tags have a direct effect on similar apps and influence both:

• The apps that point towards your app
• The apps that your app points towards to

Additional Q&A

• How is retention defined? Active users, installs/uninstalls?
• Retention ML mentioned is how many people have the app install D1/D7/D10 but Google also measures the activity of users and the number of times people open your app. If people never open your app again it will have a bad effect on your app's ranking.