If you've been doing ASO the same way since 2023, you're leaving downloads on the table. Apple and Google have both updated how they weigh keyword signals, and the strategies that worked 18 months ago are producing noticeably worse results today.
We analyzed over 2.4 million tracked keywords across the Marteso platform to build this framework. Here's what the data actually says.
The volume trap
Most teams still start keyword research by sorting by search volume and picking the highest numbers. This is the single most common mistake we see. High-volume keywords are dominated by apps with hundreds of thousands of ratings. Unless you're already in the top 10, you won't rank — no matter how well you optimize.
The better metric is the difficulty-to-volume ratio. You want keywords with enough searches to matter, but where the competition hasn't fully saturated the top results.
The framework: 3 keyword tiers
We recommend building your keyword strategy in three tiers:
- Anchor keywords (5–10): Your core category terms. High volume, high difficulty. You probably won't rank #1, but being in the top 20 still drives discovery.
- Growth keywords (20–40): Medium volume, medium difficulty. These are your workhorse terms — the ones that will drive most of your organic installs within 90 days.
- Long-tail keywords (50–100): Low volume, low competition. These punch above their weight because conversion rates are much higher when intent is specific.
Signals the algorithm now weights heavily
Based on our data from Q4 2025 through Q1 2026, the App Store algorithm shows significantly stronger correlation between installs and ranking for the following:
- Keyword appearance in reviews (not just metadata)
- Session length and retention for users who found the app via that keyword
- Velocity of installs in the first 72 hours after a metadata update
- Ratings from users acquired via organic search (vs. paid)
How to validate before committing
Never update your metadata with an untested keyword set. Use Marteso's keyword simulator to estimate rank potential before publishing. Run 2-week sprints: update, track rank changes, roll back or double down based on data.
The teams seeing the best results are running keyword updates every 3–4 weeks with tight feedback loops. Not quarterly. Not annually. Monthly.
The one thing to do today
Pull your current keyword list and remove anything with a difficulty score above 80 where you're not already ranking in the top 15. Redirect that character budget to 5–8 growth-tier keywords you haven't tested yet. Check back in two weeks.