A hundred downloads can come from a small, interested audience or a much larger audience that mostly passes by. The total alone won’t tell you which happened or whether you should rewrite a screenshot caption.
ASO metrics measure how people find your app on the App Store and download it. Reading the numbers together helps you decide where to look next: how widely the app appeared, who saw it and whether they downloaded. This guide explains the main measurements, compares two fictional reports and gives you a baseline worksheet for your next listing change.
If you’ve prepared the storyboard from our App Store screenshots guide, keep it nearby. The question now is how you’ll judge what happens after those images go live.
Give each metric a precise meaning
These short definitions follow Apple’s current metric reference:
| Metric | What it counts |
|---|---|
| Impressions | App views lasting more than one second on the App Store’s Today, Games, Apps and Search tabs; includes product page views. |
| Impressions (Unique Devices) | Devices that viewed the app, including its product page. |
| Product Page Views | Views of the listing, including pages loaded through StoreKit. |
| Product Page Views (Unique Devices) | Devices that viewed the listing. |
| First Time Downloads | First-time app downloads. |
| Redownloads | Manual redownloads; excludes automatic updates and device restores. |
| Total Downloads | First-time downloads plus redownloads. |
Apple’s Conversion Rate divides total downloads and pre-orders by unique device impressions. A pre-order counts when placed, without counting again when downloaded. Apple: Metric definitions.
Write the full metric name in your notes. “Views” is too vague: your teammate might use it for impressions while you mean product page views. Also record whether the number counts repeated activity or unique devices.
Someone can download directly from search results without opening the product page. Apple’s Page Type dimension distinguishes downloads from product pages, store sheets, events and other surfaces, including search. Apple: Filters and dimensions.
Keep that route in mind when reading the numbers. A drop in downloads doesn’t necessarily mean visitors opened the product page and decided against the app.
Write the denominator beside every percentage
For a released app with no pre-orders, the standard calculation is:
Apple Conversion Rate = Total Downloads ÷ Impressions (Unique Devices) × 100.
Suppose your report has 100 total downloads and 2,000 unique device impressions. That gives 5%. For an app with pre-orders, follow Apple’s pre-order treatment rather than adding subsequent downloads of those same pre-orders again. Apple: Analytics dashboard.
You might calculate another ratio for your own review:
Custom ratio = First Time Downloads ÷ Product Page Views × 100.
Label it with that exact formula. Both parts differ from Apple’s standard rate, and the calculation doesn’t track individual page visitors through to download. Downloads from other surfaces can enter the numerator, so the result doesn’t tell you what percentage of page visitors downloaded.
For example, 90 first-time downloads divided by 300 page views gives 30%. You can reproduce that calculation, but you cannot call it Apple’s Conversion Rate or assume those 90 downloads all came after one of those 300 views. Check the report’s available page and source breakdowns before interpreting it.
Compare two reports with the same download total
The following numbers are invented to demonstrate the reasoning. They are not Daynook results, ASO Analytics screenshots or benchmarks. Both reports cover the US territory, iPhone, App Store Search, October 1–7, 2026, with no pre-orders. Each compares with September 24–30 using the same filters.
The unique impression figures are period-level totals from the reports, not sums of daily unique counts. A device appearing on several days must not be assumed to be a different device each day.
Report A: wider exposure, slower download growth
| Measurement | Previous → current |
|---|---|
| Unique device impressions | 4,000 → 10,000 |
| Product page views | 600 → 1,500 |
| First-time downloads | 72 → 90 |
| Redownloads | 8 → 10 |
| Total downloads | 80 → 100 |
| Apple Conversion Rate | 2% → 1% |
Exposure grew by 150%, while total downloads grew by 25%. The rate fell because exposure grew faster. That alone tells you nothing about whether the screenshots got worse.
The first question is where the extra exposure came from. Did the app appear for broader queries? Did an advertising change bring a different audience? Is the result concentrated in a few days? Review source and campaign context before editing the listing.
For this fictional report, the developer should check the week’s search and advertising changes, then ask whether the listing answers the needs of the newly reached audience. If it doesn’t, they have a reason to test different wording or images. Without that evidence, rewriting all the screenshots would be a guess.
Report B: unchanged exposure, more downloads
| Measurement | Previous → current |
|---|---|
| Unique device impressions | 2,000 → 2,000 |
| Product page views | 300 → 300 |
| First-time downloads | 45 → 90 |
| Redownloads | 5 → 10 |
| Total downloads | 50 → 100 |
| Apple Conversion Rate | 2.5% → 5% |
The rate doubled with the same exposure count. Before replacing the listing, find out what changed. Different people may have seen the app, or another change may explain the improvement; the report doesn’t isolate the effect of the images.
If the developer wants more new downloads, they could revisit the keyword shortlist and check a small set of relevant queries. Could the app reach more people who need it? The report gives them a reason to investigate, but cannot identify which keyword supplied the downloads.
Report B’s 5% is not a target for Report A. Even with matching headline filters, the audiences and queries can differ. Use each app’s history and the changes it experienced to guide the investigation.
Keep source and market filters consistent
A global total can hide a drop in the market you just localized. A rate that combines all sources can change when the mix of traffic changes, even if performance within each source stays stable.
Start with one question and one manageable scope. For example: “After the English screenshot update, did US iPhone search performance change?” Within each report, use the same date range, territory, device and source definitions for every metric. The baseline and comparison cover separate periods with matching filters. If a filter isn’t available for one metric, record that limitation rather than treating the results as identical populations.
To collect the inputs in App Store Connect:
- In Apps, select your app and open Analytics.
- Under Acquisition, choose Sources. Apple says this view initially shows Product Page Views (Unique Devices).
- Use the metric dropdown to choose the measurement you need. Record its full name and unique-device setting. The worked examples above use Product Page Views, not its unique-device variant.
- Select the date range and the available territory, device and source filters. Save those choices beside the value in your worksheet.
- Repeat for the other metrics and the comparison period, checking that the scope still matches. Record any filter you cannot apply consistently.
This is an interactive review of the dashboard. For bulk source-data exports, Apple directs developers to the Analytics Reports API; record the report type and time granularity alongside any downloaded data. Apple: Acquisition.
Keep advertising in the change log. Apple says its ads performance appears in App Store search reporting, so “App Store Search” does not by itself mean an organic-only result. Apple: Monitor search performance.
The fictional reports don’t include campaign history. Check it before treating their search exposure as unpaid traffic.
Compare complete periods and record what changed
Use periods of the same length with comparable weekday coverage. A seven-day window can be convenient, but it isn’t a universal test duration. A seasonal app, a launch spike or low traffic may need a different review plan.
For the fictional screenshot project, you could record this plan before release:
- Baseline: September 28–October 4, 2026.
- Screenshot release: planned October 5; replace with the actual live date.
- Comparison: October 5–11, if the assets were live throughout it.
- Scope: US territory, iPhone, App Store Search.
- Question: did exposure, first-time downloads and the standard rate move together?
- Review condition: both exports cover complete periods with the same definitions and filters.
This is a fictional review plan. The dates don’t establish that you have enough data to judge the change. If release slips or the comparison includes partial reporting, move the window. Save the screenshots and metadata that were live in each period.
Record product releases, price changes, promotions, outages and advertising alongside the listing change. If several changed together, the before-and-after report cannot isolate a caption’s effect. A controlled creative experiment needs its own setup and interpretation; keep its results separate from this general dashboard comparison.
Investigate missing data before calling it a decline
An empty chart and a zero are different observations. Check the selected dates, filters and report availability before concluding that activity stopped.
For a newly launched app, check whether the relevant metrics are available yet. Apple lists at least five first-time downloads or pre-orders for App Store metrics, and at least five first-time downloads for download metrics. These are metric-availability requirements, separate from usage-data opt-in and privacy thresholds for individual report rows. Apple: Metric availability.
Apple’s Analytics dashboard applies privacy and availability conditions, and its usage data covers people who agreed to share it. Usage figures therefore don’t represent every download. Apple: Dashboard privacy and data availability.
If you use the Analytics Reports API, distinguish the first report’s arrival from its data being complete. An initial ongoing request produces its first report approximately 24–48 hours later. That does not mean every recent reporting day is finalized. Apple: Analytics Reports API.
Apple’s API overview gives a two-day completeness window, but the detailed documentation specifies the window for each report: App Store Downloads lists two days, while Discovery and Engagement lists three. Check the Completeness field for every report you compare. Late-arriving events can appear in later instances; use Apple’s data completeness and corrections guidance when downloading them.
Some reports also apply privacy safeguards that omit low-volume rows; the rules depend on the report and whether it is standard or detailed. Apple: Report privacy. Dashboard and third-party refresh schedules need to be checked separately.
Mark missing or incomplete values in your worksheet instead of entering zero. Return to the comparison when both periods have comparable data. Until then, you don’t have a conclusion.
Your baseline measurement sheet
Copy this before the next listing change:
- App and owner:
- Question being investigated:
- Baseline and comparison dates:
- Export time and reporting source:
- Territory, device, acquisition source and any page filters:
- Exact metric names, including unique-device settings:
- Impressions and product page views:
- First-time downloads, redownloads and total downloads:
- Apple Conversion Rate as reported:
- Any custom ratio, with its full formula:
- Data completeness or missing values:
- Listing assets and actual release date:
- Other product, pricing and campaign changes:
- Observation, possible explanations and next check:
Keep the observation separate from the explanation. “Impressions rose and the rate fell” can be verified from Report A. “The new caption attracted the wrong users” needs more evidence.
For a screenshot change, use the worksheet to choose your next step: check who saw the app, review what the images promise or wait for more data.
Frequently asked questions
What are the most useful ASO metrics to start with?
Start with exposure, first-time downloads and a clearly defined conversion rate. Add page views and source breakdowns to investigate what changed. Choose the scope and question first so you don’t collect numbers you won’t use.
Are impressions and product page views the same thing?
No. Impressions count app views across the App Store’s Today, Games, Apps and Search tabs and include product-page views. Product page views count visits to the listing, including through StoreKit. Both have unique-device variants. A person can download directly from search without visiting the product page, so the two counts are not mandatory steps in a funnel.
Can I divide downloads by page views to get conversion rate?
Yes, as a custom ratio with its numerator and denominator clearly labeled. Apple’s standard Conversion Rate uses total downloads and pre-orders divided by unique device impressions, with a pre-order counted once. Downloads divided by page views is a different calculation. Aggregate counts alone do not show which page visitors downloaded, so the ratio is not proof of visitor-to-download conversion.
What is a good App Store conversion rate?
There’s no single target that works for every app. Compare the same scope over time and check what changed in the audience, product and campaigns. A higher rate won’t necessarily bring more downloads if the app reaches far fewer relevant people.
Why did downloads rise while conversion fell?
Exposure may have grown faster than downloads, as in Report A. Inspect the traffic behind the increase before blaming a listing change. The counts tell you what moved; they don’t identify the cause.
Does better keyword ranking prove an ASO change worked?
A keyword rank tells you where the app appeared for a particular query and storefront. It doesn’t tell you which downloads came from that query or whether conversion improved. Record the ranks and dates beside your acquisition report and look for evidence connecting them.
Choose the next check from the evidence
Save one report with its filters and write a sentence describing what changed. Then name what you need to check before editing the listing. For Report A, that means finding out where the extra exposure came from.
Next, track keyword ranking trends without mistaking a brief change or a missing check for a product problem. You can follow keyword positions and history in ASO Analytics; use App Store Connect for the source breakdowns and standard conversion reporting discussed here.
Preparing a release? Use the free ASO pre-release checklist to review your listing and measurement plan.