Do apps climb because competitors leave the chart?

Big app-chart climbs usually involve overtaking competitors.

We tested whether apps often climb because competitors disappear from the top 100. Across 6,819 large climbs, that was rarely the main explanation.

Prepared 27 October 2026 · Source observations: 19 September to 26 October 2026, UTC · Apple App Store and Google Play · Top-free app charts

After subtracting new arrivals above an app, departures explained at least half its gain in just 1.33% of Apple climbs and 2.93% of Google Play climbs. Most big ranking gains involved improving position against apps that remained in the chart.

1. What we tested, in plain English

The original idea was simple: your app might jump from #50 to #30 because 20 apps ahead of it leave the top 100. Your ranking improves, but you have not passed any of the competitors that remain.

That was a hypothetical example. We checked how often something similar actually happened, using apps that climbed at least 10 positions between consecutive daily charts. Both charts had to contain a complete, valid top 100 and be recorded 18 to 30 hours apart.

There is one important adjustment. If apps leave from above you but other apps enter above you, those arrivals take positions back. We therefore count departures above the app minus new arrivals above it. We then ask whether that balance explains at least half of the climb.

One climb means one app, in one country, over one daily chart pair. The same app can appear on several dates or in several countries. These are 6,819 ranking events across 1,643 store-specific app IDs, not 6,819 independent ASO campaigns. The last column retains the original departure-only calculation so the adjustment is transparent.

Even before subtracting new arrivals, the departure-only figure was just 6.64% for Apple and 7.59% for Google Play. The result does not depend on introducing the stricter adjustment.

2. Where the gained positions came from

For each climb, we separated movement against continuing competitors from the balance of apps entering and leaving above it.

On Apple, a large climb averaged about 20.6 positions. Roughly 19.0 came from improved standing against apps still in the chart, and 1.5 from departures minus arrivals. Google Play showed the same broad pattern.

The final column divides the total contribution from continuing competitors by all positions gained in the selected climbs. It is a share of chart positions, not a share of downloads, users or ASO success. Displayed averages are rounded.

3. A real example: Disney+ climbed 20 places

On Apple’s US chart, Disney+ moved from #77 to #57 between 10 and 11 October.

The two departures helped, but the two arrivals cancelled that help. The full 20-position gain remained when we compared Disney+ only with apps visible on both days.

This is evidence of improved relative chart position. It does not tell us why Disney+ improved or whether an ASO change contributed.

Four real climbs, with very different explanations

Indeed’s US Google Play climb was mostly movement against continuing competitors: 21 of its 22 gained positions. eBay’s German Apple climb was an exception: departures minus arrivals accounted for 10 of its 15 positions.

Examples were selected after the analysis to explain the arithmetic and show exceptions. They are not a random sample. Full timestamps, app IDs, source snapshot IDs and the identity of every contributing competitor are included in the evidence.

4. The strongest version of the theory did happen, but only three times

Across all 6,819 large climbs, we found three where an app gained positions overall while making no net improvement against competitors that remained in the chart. All three were on Google Play in the earlier period.

In India, it climbed from #87 to #77 on 29 September. It passed four continuing competitors but was passed by five others, a net loss of one position against that group. Seventeen previously higher apps left the chart, while six new arrivals appeared above it. The arithmetic was 4 + 17 − 5 − 6 = 10 positions gained.

Even these three cases included some direct overtakes. We found no qualifying climb where the app passed zero continuing competitors. The initial hypothetical example therefore did not occur literally in this sample.

These exceptions show that chart entries and exits can matter. They do not support treating most large ranking gains as a result of disappearing competitors.

5. The result repeated in the later period

In the later period, the share fell to 0.89% on Apple and 2.12% on Google Play. The second uncertainty estimate resamples groups of neighbouring dates, since the same chart change can affect many apps at once.

These intervals describe variation within the observed sample. They do not cover every possible collector error or establish how all apps behave. The date split repeats the test within an already-used dataset; it is not a new external sample.

Changing the rules did not rescue the theory

For climbs of 20 positions or more in the later period, there were zero cases where entries and exits explained at least half the gain: 0 of 779 Apple climbs and 0 of 426 Google Play climbs.

This zero is specific to that later sample. In the earlier period, there were 7 such Apple cases and 23 Google Play cases. It is not evidence that the pattern never happens.

Giving each app equal weight instead of each event also left the later result small: 0.81% on Apple and 2.43% on Google Play.

6. What this means for ASO

A large chart climb usually represented improved standing against continuing competitors in this sample. Dismissing those gains as mostly the result of competitors disappearing would have been wrong.

The finding is about the meaning of a ranking change. It does not prove that the app gained downloads or that ASO caused the improvement. We did not have keyword impressions, listing conversion, download counts, advertising exposure or a controlled ASO test for these apps.

These are top-free app charts. They are not keyword search results. The conclusion should not be presented as a finding about keyword-ranking gains.

7. Method, prediction and verification

1. Validate the chart source. We used 714 source snapshots marked valid and complete for the top-free, all-apps chart. We independently checked that every snapshot had 100 unique app IDs and the ranks 1 through 100. All 714 passed.
2. Build comparable daily charts. Select the latest eligible chart for each store, country and UTC date. This produced 671 daily charts. Of these, 57 had no chart on the preceding calendar date and 18 more failed the 18-to-30-hour timing rule. The remaining 596 daily chart pairs formed the sample.
3. Track apps visible at both ends. There were 54,973 app comparisons. We checked the ranking identity for every one, then selected the 6,819 gains of at least 10 positions. Apps entering from outside the top 100 were not assigned an invented starting rank.
4. Separate four types of movement. Add continuing competitors passed and previously higher apps that departed. Subtract continuing competitors that passed the focal app and new arrivals above it. These four quantities exactly reproduce the rank change.
5. Repeat and vary the test. Analyse the stores and time periods separately; tighten timing; increase the required gain; restrict starting ranks; keep only one climb per app per period; exclude the largest chart turnovers; and use local dates.

What the data cannot settle

• A departure is an observed absence from the next top 100. The app might have fallen below #100, become unavailable, or been affected by a collection issue. We do not know which.
• The accounting is descriptive. Removing apps from a calculation is not a real-world experiment in which those apps vanished and everything else stayed fixed.
• The sample is limited. It covers one third-party collector, 38 UTC dates and the markets observed in that source. It excludes incomplete charts and large gaps. Selection can affect the result.
• Overtaking does not prove download growth. Relative position cannot tell us whether the focal app’s downloads rose, fell or stayed flat. Apple documents impressions, downloads and conversion as separately measured metrics; this study does not contain those private performance measures.
• There is no ASO attribution. We cannot assign movement to keywords, screenshots, ratings, advertising or other causes from this chart data alone.

8. Download the evidence

The report includes a downloadable evidence ZIP with the source chart entries, snapshot index, metadata used for names, full event tables, competitor IDs, protocol, analysis code, verification code and results.

Evidence ZIP: 10.2 MB. The ZIP is embedded in this HTML. Python and NumPy are required to rerun the calculations. No images or external scripts are required to read the report.

1. AppOrbit pinned dataset
   Previously preserved chart entries, snapshot index, metadata extract, format and license. Metadata supplies names only; all ranking calculations use chart records.
2. Apple: Analytics metric definitions
   Context for the distinction between chart position and directly measured impressions, downloads and conversion. These private metrics were not available for this study.

Publication finding: Across 6,819 large daily app-chart climbs, competitors leaving the top 100 were rarely the main explanation once new arrivals were counted. Most climbs reflected improved relative standing against apps that remained visible. That makes the ranking gain meaningful as a relative position change, while leaving its cause and download impact unresolved.

Prepared for Gabriel Machuret · ASO Experiment 8 · Source attribution: AppOrbit. Database rights under ODbL 1.0; third-party store content has separate rights. Analysis dates are UTC unless stated otherwise.