If you have ever put your ad platform's install count next to App Store Connect and found they disagree, nothing is broken. Those numbers are not measuring the same thing, and they never will.
Understanding why saves a lot of wasted investigation.
What each source actually measures
Ad platforms report installs they can attribute to an ad, within their own attribution window, using their own rules. They have a structural incentive to claim credit, and each platform's rules differ.
Store consoles (App Store Connect, Google Play Console) report actual downloads, with no idea what drove them. This is the closest thing to ground truth for volume, and it tells you nothing about source.
Link analytics report clicks and scans — how many people engaged with your link, from where and on what device. This is precise, because it happens on infrastructure you control. It stops at the store boundary.
Product analytics report first app opens, which is not the same as installs. People download and never open. Others open days later.
Four sources, four definitions. Disagreement is the expected state.
The gap that cannot be closed
Between "clicked your link" and "installed the app" sits the app store, and both Apple and Google have deliberately made that boundary opaque.
Historically, various techniques inferred the connection. Most have been closed off. Apple's App Tracking Transparency and SKAdNetwork, and Google's Privacy Sandbox work, replaced deterministic user-level tracking with aggregate, privacy-preserving reporting — delayed, sometimes noise-added, and not resolvable to individuals.
This is a permanent change, not a temporary inconvenience. Attribution today means directional confidence, not certainty.
How to reason about it instead
Compare like with like over time. Your click numbers versus your click numbers, week over week, is a valid comparison. Your click numbers versus a platform's attributed installs is not.
Watch ratios, not absolutes. If clicks are flat and store downloads drop, something changed at the store — a listing problem, a review issue, a ranking shift. If clicks drop and downloads follow, the problem is upstream in your channels.
Instrument what you control. You own the click. You own the scan. You own the source tag. That data is complete and accurate, and it is enough to compare channels against each other, which is usually the actual decision you are trying to make.
Accept overlap. Two platforms both claiming the same install is normal, not fraud. Summing attributed installs across platforms will overcount, every time.
Practical setup
Tag every channel separately. Same destination, distinct source tags for Instagram bio, TikTok bio, newsletter, podcast, paid social. Without this, all traffic collapses into one undifferentiated bucket.
Use one link per campaign, not per platform. Device routing handles iOS versus Android. Splitting by platform doubles your links and halves your clarity.
Record the referrer. Knowing traffic came from a specific partner or publisher is often more actionable than knowing the raw total.
Check store-side numbers weekly. They are the volume ground truth even though they carry no source information.
Write down your definitions. "Installs" meaning attributed installs in one report and downloads in another is how teams end up arguing about a discrepancy that was never real.
What good looks like
You should be able to answer:
- How many people clicked or scanned our install link this week?
- Which channel produced the most?
- Did store downloads move in the same direction?
- Which campaign had the best click-to-download ratio?
You will not be able to answer "exactly which of these specific people installed," and any tool promising that deserves scrutiny.
How Bridgly fits
Bridgly measures the part that is genuinely knowable: clicks and scans, by source, channel, device, and referrer, on infrastructure you control.
Referrers are first-class, so partner and creator traffic can be separated without creating a new link for each one. Install attribution via store sync is available on Core and Pro, which closes part of the gap where the platforms allow it.
Start free · Read: one link in bio for iOS and Android
The short version
- Four sources, four definitions — they will never match.
- The store boundary is intentionally opaque, permanently.
- Compare like with like, and watch ratios over time.
- Instrument clicks and scans, which you fully control.
- Expect cross-platform overcounting when summing attributed installs.
Frequently asked questions
- Why do my ad platform and App Store Connect report different install numbers?
- They measure different events over different windows. Ad platforms report attributed installs credited to an ad within their attribution window, using their own rules. Store consoles report actual downloads regardless of source. Overlap is partial by definition, and the two numbers should never be expected to match.
- What is an attribution window?
- The period after a click or view during which an install is credited to that touchpoint. Windows differ by platform and are configurable, so the same install can be attributed by one platform and not another depending on how long ago the interaction happened.
- Can you track exactly who installed an app from a link?
- Not deterministically in most cases. App stores deliberately break the chain between a web click and an install for privacy reasons, and both Apple and Google have tightened this substantially. Modern attribution relies on privacy-preserving aggregate frameworks and probabilistic signals, which means directional confidence rather than certainty.
Read more
Attribution · 1 min read
UTM parameters for app campaigns: a naming convention that survives
UTM tags are trivial to add and easy to ruin. Here is a naming convention that still reads clearly after a hundred campaigns, and the specific mistakes that make app campaign data useless.
Attribution · 2 min read
Branch vs AppsFlyer vs a link platform: which do you actually need
Branch and AppsFlyer are mobile measurement partners built for multi-network paid acquisition. Most teams comparing them do not need either yet. Here is where the line falls and what the cheaper answer looks like.
Fundamentals · 1 min read
Bitly vs Dub vs Bridgly: picking by job
Bitly shortens. Dub pays digital partners. Bridgly routes app installs and ties physical QR scans to attributed sales and commissions.