Ranking #3 for a keyword and ranking #30 for the same keyword are not two points on the same line. They are two different businesses. One gets discovered every day. The other technically exists.
This guide covers how App Store search ranking works at a high level, how to pick keywords actually worth tracking, what a rank number really means (less than you think), and a weekly routine that takes about ten minutes. Written for indie devs without an ASO agency on retainer — which is most of us.
Why rank is the whole game
Apple's own Apple Ads marketing pages have long cited that roughly 65% of App Store downloads happen directly after a search. Take the exact figure with the usual grain of salt — it comes from a page selling search ads — but nobody who does ASO seriously disputes the direction: search is the front door of the App Store, and for most indie apps it is the single biggest discovery channel that exists.
Search traffic also decays brutally with position. On a phone screen, the first result takes most of the taps, the top three take nearly all of them, and anything past the first scroll might as well be unlisted. Rank 3 versus rank 30 is not "10x fewer installs." It is closer to something versus nothing.
Indies feel this harder than anyone. You have no brand searches. You probably have no featuring contact at Apple, and editorial placement is a lottery ticket. Keyword rank is the one discovery lever you can actually observe, reason about, and move.
How App Store search ranking works, roughly
Apple does not publish the algorithm. Everything below is the practitioner consensus, built from Apple's public documentation plus years of collective observation. Treat it as a working model, not gospel.
- Text relevance. Your app name carries the most weight, the subtitle carries meaningful weight, and the hidden 100-character keyword field fills in the rest. Apple's own search documentation confirms these metadata fields feed the index. The long description is widely believed not to be indexed for iOS search — write it for humans, not the algorithm.
- Ratings: count and velocity. An app with 4.7 stars and a steady stream of fresh ratings outranks a near-identical app with 4.2 and a dry spell. Recency appears to matter, not just the lifetime total.
- Downloads and conversion for the query. If people search a term, tap your app, and install it, Apple learns your app satisfies that query. This is the flywheel: rank drives installs, installs reinforce rank. It is also why breaking into a competitive term is slow and losing a position can compound.
The practical takeaway: metadata gets you considered, conversion keeps you there. You cannot observe Apple's internals, but you can observe the output — your rank, per keyword, over time. That is the entire point of tracking.
Picking keywords worth tracking: relevance × volume × winnability
The classic indie mistake is dumping 50 keywords into a tracker and drowning in a wall of numbers nobody acts on. Fifty tracked keywords is zero tracked keywords. Pick 3 to 5 and actually watch them.
Score candidates on three axes:
- Relevance. Would someone typing this phrase be genuinely happy to land on your app? If the honest answer is "sort of," ranking for it produces taps that do not convert — which, per the model above, can hurt you.
- Volume. Are real humans typing this? You can gauge this roughly from App Store autocomplete suggestions and from popularity scores in Apple Ads keyword planning. Precision does not matter; order of magnitude does.
- Winnability. Look at who currently holds the top ten. If it is wall-to-wall apps with tens of thousands of ratings, a generic head term is not your fight yet. A two-to-three-word phrase where the incumbents look beatable is.
A hypothetical example, for a solo-built interval timer app:
| KEYWORD | RELEVANCE | VOLUME | WINNABILITY | TRACK? |
|---|---|---|---|---|
| timer | Medium | Huge | None — giants everywhere | No |
| interval timer | High | Good | Contested but plausible | Yes |
| hiit workout timer | High | Moderate | Beatable incumbents | Yes |
| tabata timer | High | Moderate | Beatable incumbents | Yes |
| fitness app | Low | Huge | None | No |
Numbers and verdicts above are illustrative, not measured. The shape is the point: skip the head terms you cannot win and the vague terms you should not want, then track the handful in the middle where movement is both possible and meaningful.
What "rank" actually means
Before you obsess over a number, understand what the number is:
- Rank is per storefront. Your position for "habit tracker" in the US store says nothing about Germany or Japan. Each country is a separate index with separate competitors. Track the storefronts where your revenue actually lives.
- Results are somewhat personalized and inconsistent. Two people searching the same term at the same time can see slightly different orderings — device, prior behavior, and index rollout timing all appear to nudge results. Your friend's screenshot is anecdote, not data.
- Rank fluctuates constantly. Positions wobble day to day and sometimes hour to hour without anyone doing anything. A single reading is noise.
Track trends, not readings. "Slid from ~4 to ~9 over two weeks" is a signal worth investigating. "I was #4 yesterday and #6 today" is weather. Any single rank check is one noisy sample from a distribution — the moving average is what you manage.
Three ways to actually track it
Manual incognito searches
The zero-cost option: type each keyword into App Store search, count down to your icon, write it in a note. It works, badly. Personalization pollutes the results even on a "clean" device, you cannot easily check other storefronts, there is no history unless you are disciplined about the spreadsheet, and after week three you will quietly stop. Manual tracking mostly measures your own consistency, and indies lose that bet.
Web SaaS dashboards
The established route: browser-based ASO platforms that track keyword positions across storefronts, usually alongside download estimates and competitor metadata. They are genuinely capable — most are built for studios and agencies managing many apps, priced accordingly, and involve creating an account and keeping your keyword strategy on someone else's servers. If you are running ASO across a portfolio, this tier makes sense. We compare the landscape in our ASO tools guide for indie developers.
On-device tracking
The newer option, and the one Rival Radar takes: an iOS app that queries public App Store data directly from your device. You pick your keywords once, and each scan rechecks where you rank and where your named rival ranks for the same terms, per storefront, side by side. The you-versus-them framing matters — rank is a relative game, and "we're #7" reads completely differently depending on whether your rival is #3 or #15. History stays on your device: local-first, no account, and keyword tracking runs alongside the same scans that watch rivals' listing changes and reviews.
None of these methods is wrong. The right one is whichever you will still be using in month three — a tracker you abandon has an ROI of zero.
When your rank drops: debug it like a bug
A real drop — trend, not wobble — has a cause. Work the checklist in order:
- Confirm it is a trend. Several consecutive readings moving one direction, or one bad sample? If it is one sample, close the tab and go ship something.
- Check your rivals' listings first. The most common cause of "we dropped" is "they moved." A competitor putting your keyword in their title or subtitle can displace you without you changing a thing. This is why Rival Radar pairs rank tracking with listing-change detection: when a rank dips, you can check whether a rival shipped a metadata change in the same window and see the correlation instead of guessing. Our guide to tracking competitor apps on the App Store covers this angle in depth.
- Check your rating velocity. A recent one-star cluster or a stretch with no new ratings can drag rank. If reviews are the problem, mining reviews systematically tells you what to fix.
- Check what you changed. A metadata edit, a new release, an icon swap — anything that shipped near the drop is a suspect. You would not debug a crash without reading the diff; same discipline here.
- Consider seasonality. Fitness terms spike in January, study terms in September. Sometimes your rank fell because the query changed, not because you did anything wrong.
The weekly 10-minute routine
Rank tracking pays off through boring consistency, not heroic deep-dives. Once a week:
- Minutes 0–3: Scan your 3–5 keywords. For each: trending up, flat, or down over the last few weeks? Ignore single-day wiggles.
- Minutes 3–6: For anything trending down, run the debug checklist above — rival listing changes first, then your rating velocity.
- Minutes 6–9: Note one action if one is warranted: a subtitle tweak to answer a rival's move, a review-prompt adjustment, a keyword-field swap for your next release. One change at a time, or you will never know what worked.
- Minutes 9–10: Log the week in one line. "All flat, no action" is a perfectly good entry.
Change one metadata variable per release, then watch the trend for two to three weeks before judging. Shotgunning five changes at once means you learn nothing regardless of the outcome.
That is the whole practice: a handful of honest keywords, trends over readings, rivals watched alongside your own numbers, and ten disciplined minutes a week. If you want the broader pre-flight list this slots into, our ASO checklist for indie iOS devs is the companion piece. Now close the dashboard and get back to the build.
Rival Radar turns public App Store data into competitive intel — change detection, keyword ranks, review mining. Local-first, no account. Free tier included.
GET RIVAL RADAR →