Most teams find out a deep link is broken when a campaign underperforms and someone finally taps the link by hand. By then the spend is gone. Deep link health is measurable, and the metrics that show it are not the ones most dashboards surface.
The problem is that the obvious number, click count, tells you a link was tapped, not that it routed anyone to the right place. A link can rack up thousands of clicks while quietly dropping every user on the home screen. This post covers the metrics that actually reveal routing health, the diagnostic cuts that catch breakage early, and the simple monitoring view that keeps it visible.
What Deep Link Analytics Measures
Deep link analytics: the measurement of whether a deep link routed users to the intended destination, tracked through metrics such as match rate, deferred deep link success rate, time-to-route, and fallback rate, as distinct from the attribution credit the same link carries.
A single link answers two different questions, and teams often conflate them:
-
Did routing work? Did the user land on the exact screen the link promised? That is a health question.
-
Who gets credit? Which campaign drove the resulting install or action? That is an attribution question.
Both matter, but a healthy attribution number can sit on top of badly broken routing. A click count alone tells you nothing about either. For the foundational concepts, our complete guide to deep links is the primer; this post is about measuring them once they are live.
The Core Deep Link Health Metrics
Four metrics give you a near-complete picture of routing health:
-
Match rate. The share of clicks that resolved to the intended in-app route. This is the headline health metric. A falling match rate is the first sign something is broken.
-
Deferred deep link success rate. For users who had to install first, the share that landed on the routed screen rather than the home screen. This is where most silent failures hide, because the user journey spans an install.
-
Time-to-route. The latency from tap to the correct screen appearing. Slow routing reads as a broken experience even when the destination is right.
-
Fallback rate. The share of taps that fell back to the app store or a web page instead of routing. A rising fallback rate is leaked intent.
Together these separate "the link was tapped" from "the link worked", which is the whole point. Our overview of how deep linking drives higher conversion rates explains why each percentage point of match rate maps to real revenue.
The Diagnostic Cuts That Catch Breakage Early
A blended number hides problems. Segment these four metrics to find what a single average masks:
-
By platform (iOS vs Android). Divergence here is the most common first red flag. One platform collapsing while the other holds usually points to a platform-specific configuration issue.
-
By source or channel. If one channel's match rate craters, the link template for that channel is almost always malformed.
-
By OS version. Deferred success dropping on a specific OS version catches regressions like the ones recent Android and iOS releases introduced.
-
By in-app browser. Isolate taps inside Instagram, TikTok, and other webviews, where universal links behave differently and routing often fails quietly.
Across deep linking onboardings, the most common silent failure we see is a platform-specific match-rate gap, healthy on Android, collapsing on iOS or the reverse, that goes unnoticed for weeks because the blended match rate still looks acceptable. Segmenting by platform is what surfaces it. The 10 non-negotiable deep linking features checklist covers the analytics and export capabilities you need to run these cuts at all.
How to Build a Deep Link Health View
You do not need a complex dashboard. You need four metrics in one place and an alert.
-
Put the four metrics on one screen. Match rate, deferred success, time-to-route, and fallback rate, each segmented by platform.
-
Set threshold alerts, not manual checks. Trigger a notification when match rate or deferred success drops below your healthy band, so you find out the day it breaks, not at the campaign post-mortem.
-
Run a weekly five-minute review. Scan the segmented view, confirm no platform or channel has diverged, and move on.
The aim is to make routing health a passive signal you are alerted to, not an active investigation you remember to run. Pre-launch, the deep link QA checklist prevents the obvious breaks; this view catches the ones that appear after launch.
What Good Looks Like and What to Action
Healthy bands vary by app and channel mix, but useful rules of thumb:
-
Match rate on owned channels should sit comfortably high, typically well above the rate you see on hostile in-app browser traffic. A sudden drop of more than a few points is worth investigating the same day.
-
Deferred success is naturally lower than direct match rate because the journey crosses an install, but it should be stable. A steady decline usually means an install-referrer or routing-sequence regression.
When a metric dips, triage by likely cause in order:
-
A single channel down points to a malformed link template.
-
One platform down points to a configuration or OS-version issue.
-
In-app browser segment down points to a webview routing quirk, not a real outage.
Knowing which bucket you are in tells you whether to fix a campaign template yourself or escalate to engineering.
How This Looks Inside an MMP
Deep link analytics and attribution should not live in separate tools, because a broken route and a missing conversion are often the same incident viewed from two angles. When routing health and attribution sit together, a dropping match rate and the revenue dip it causes show up side by side.
Platforms like Linkrunner keep routing health and attribution in one view, so you can spot a falling match rate without exporting a routing report and an attribution report and reconciling them by hand. The case for keeping them unified is laid out in why deep linking and attribution should never be separate products, and the deferred deep linking and webhooks docs show how the health signals are surfaced and exported.
FAQ
What is a good deep link match rate?
It varies by channel, but owned channels (email, SMS, your own QR codes) should sit high, while hostile in-app browser traffic runs lower. The signal that matters most is a sudden drop from your own baseline, not a universal target number.
How is deep link analytics different from attribution?
Deep link analytics measures whether routing worked, did the user reach the promised screen. Attribution measures who gets credit for the resulting install or action. A link can attribute correctly while routing badly.
How do I measure deferred deep link success?
Track the share of users who installed first and then landed on the routed screen rather than the home screen. It is naturally lower than direct match rate, so watch the trend rather than the absolute figure.
Why do my deep links work on one platform but not the other?
Platform-specific configuration is the usual cause, AASA or universal link setup on iOS, App Links or install referrer on Android. Segmenting match rate by platform exposes the gap a blended number hides.
Can I get alerts when a deep link starts failing?
Yes. Set threshold alerts on match rate and deferred success so you are notified when either drops below your healthy band, rather than discovering it during a campaign review.
Closing
Deep link health is not something you hope is fine, it is something you measure. Match rate, deferred success, time-to-route, and fallback rate, segmented by platform and channel, turn silent routing failures into a signal you catch the day they appear instead of the week the budget is already spent.
If you want routing health and attribution in a single view with alerts when a link starts slipping, request a Linkrunner demo. A practical first step is to build the four-metric view this week, segment it by platform, and set one alert on match rate, then let the dashboard tell you when something breaks.