Your app marketing numbers don’t add up because different platforms, tools, and attribution models each count installs and conversions using their own logic. An install reported by Meta may not match what your Mobile Measurement Partner (MMP) records, and neither may align perfectly with what Apple or Google reports. This is a normal part of mobile attribution, not a sign that something is broken. The sections below explain why these gaps exist and what you can do about them.
Why do different platforms report different install numbers?
Different platforms report different install numbers because each one applies its own attribution rules, counting windows, and data sources. Meta counts a conversion when a user clicks your ad and installs the app within its attribution window. Google does the same. Your MMP applies a separate set of rules to decide which platform gets credit. When two platforms both claim the same install, your total appears inflated compared to what your MMP reports.
This overlap is called double counting, and it is one of the most common sources of confusion in app marketing reporting. A user might see your ad on Meta, then search for your app on Google before downloading it. Both platforms record a conversion. Your MMP, however, assigns credit to only one source based on its attribution logic, typically last click or a probabilistic model.
On top of that, platforms use different attribution windows. Meta might claim any install within seven days of a click, while your MMP uses a three-day window. Installs that fall outside your MMP window still count on Meta’s side, widening the gap further.
What is mobile attribution and how does it work?
Mobile attribution is the process of identifying which marketing source, campaign, or ad drove a specific app install or in-app event. When a user clicks an ad and opens the app, an attribution tool records the source, matches it to the install, and assigns credit. This tells you which channels are actually driving results so you can allocate budget effectively.
In practice, attribution works through tracking links and SDKs. When a user clicks your ad, the attribution platform records a fingerprint of that interaction. When the app opens for the first time, the SDK inside the app sends a signal back to the attribution platform, which matches the install to the original click.
Most businesses use a dedicated MMP such as Adjust, AppsFlyer, or Branch to handle this process. These tools sit above all your ad platforms and apply consistent rules across every channel, giving you a single, unified view of your app install tracking data. Without an MMP, you are left comparing platform-reported numbers that will almost never agree.
What causes discrepancies between my MMP and ad platform data?
Discrepancies between your MMP and ad platform data are caused by differences in attribution windows, attribution models, time zone settings, and data processing delays. Each factor introduces a gap between what a platform claims and what your MMP records. Understanding which factor is driving your specific discrepancy helps you decide how to respond.
The most common causes include:
- Attribution window mismatches: Your MMP may use a shorter lookback window than the ad platform, so some installs the platform claims fall outside the MMP’s counting period.
- View-through attribution: Some platforms count installs from users who saw an ad but never clicked it. Most MMPs default to click-based attribution only, so these installs appear in platform data but not in your MMP.
- Time zone differences: If your MMP and your ad platform operate in different time zones, installs near midnight can shift between reporting days.
- Data processing delays: Platforms often process and finalize conversion data at different speeds. A comparison pulled on the same day may show different totals simply because one source has not finished processing.
- Re-engagement vs. new install logic: Some platforms count re-engaged users as new installs. Your MMP may classify the same user as a returning user, removing them from new install counts.
A discrepancy of 10 to 20 percent between your MMP and a platform is generally considered acceptable. Gaps larger than that are worth investigating systematically, starting with attribution window alignment.
How does Apple’s SKAdNetwork affect app marketing reporting?
Apple’s SKAdNetwork (SKAN) affects app marketing reporting by replacing device-level user data with aggregated, privacy-preserving conversion signals. Since Apple’s App Tracking Transparency (ATT) framework launched, most iOS users do not consent to tracking. For those users, SKAN is the only attribution mechanism available, and it introduces significant limitations that widen the gap between platform and MMP data.
Under SKAN, Apple sends a postback to the ad network after a delay of up to several days, with no user-level detail attached. The conversion value is limited, the timing is unpredictable, and not every install generates a postback at all. This means your iOS app install tracking data is inherently less complete than Android data, where Google’s Privacy Sandbox is still more permissive at this stage.
The practical effect is that ad platforms running iOS campaigns often report fewer conversions than they actually drove, because some SKAN postbacks are delayed, dropped, or aggregated in ways that prevent matching. Your MMP receives the same incomplete signals and tries to model the gaps, but modeled data always carries uncertainty. This is why your iOS numbers often look weaker than your Android numbers even when campaign quality is comparable.
Which data source should I trust when numbers don’t match?
When your app marketing numbers don’t match, your MMP is the most reliable single source of truth for cross-channel performance. It applies consistent attribution rules across every platform, removes duplicate conversions, and gives you a unified view of where installs actually came from. Ad platform data is useful for understanding platform-specific performance, but it should not be used for cross-channel comparison.
Think of it this way: ad platforms are built to report the value they delivered. They have a commercial incentive to show strong numbers, and their attribution logic reflects that. Your MMP has no stake in which channel wins credit. It applies the same rules to every source, which makes it the right tool for budget decisions and channel comparison.
That said, neither source is perfect. For iOS specifically, SKAN data gaps mean that even your MMP is working with incomplete information. In those cases, use your MMP data as the baseline and treat platform-reported iOS numbers as directional signals rather than exact figures. For Android, the gap between MMP and platform data is typically smaller and easier to reconcile.
How can I reduce discrepancies in my app marketing data?
You can reduce discrepancies in your app marketing data by aligning attribution windows across your MMP and ad platforms, auditing your event tracking setup, and standardizing how you compare data across sources. You will never eliminate all gaps, but a structured approach brings your numbers close enough to make confident decisions.
Start with these practical steps:
- Align attribution windows: Set the same lookback windows in your MMP and your ad platforms wherever possible. Even a one-day difference can produce meaningful gaps at scale.
- Disable view-through attribution on platforms where your MMP does not support it: If your MMP only counts click-based installs, view-through conversions on Meta or Google will always create a discrepancy. Either enable view-through in your MMP or turn it off at the platform level.
- Standardize time zones: Make sure your MMP and every ad platform report in the same time zone. UTC is a safe default.
- Give data time to process: Avoid comparing numbers in real time. Pull final reports at least 48 to 72 hours after the period closes to allow all platforms to finish processing.
- Audit your in-app event tracking: Misconfigured events create reporting gaps that compound over time. Verify that your SDK is firing the correct events at the correct moments in the user journey.
- Use your MMP as the single source of truth: Build your dashboards and budget decisions around MMP data. Use platform data to optimise within each channel, not to compare across channels.
If you are running iOS campaigns, invest time in your SKAN conversion value schema. A well-configured schema gives you more usable data within the constraints Apple allows, reducing the modeling gap your MMP has to fill.
Getting this setup right requires experience with platforms like Adjust, AppsFlyer, and Branch, as well as a clear understanding of how each ad network handles attribution. If your discrepancies are persistent or larger than expected, it is worth getting a specialist to audit your full app growth stack. At Wuzzon, we help app teams untangle exactly these kinds of reporting challenges, from attribution window alignment to SKAN configuration and in-app event audits. If your numbers are not adding up and you want a clear picture of what is actually driving your growth, talk to one of our specialists and we will help you make sense of your data.
Frequently Asked Questions
How do I know if my attribution discrepancy is a real problem or just normal variance?
A discrepancy of up to 10–20% between your MMP and an ad platform is generally within the expected range and does not require urgent action. If your gap consistently exceeds 20%, or if it suddenly widens without any change to your campaigns, that is a signal worth investigating — starting with attribution window alignment, time zone settings, and your SDK event configuration. Tracking discrepancy percentages over time rather than reacting to individual data points helps you distinguish normal variance from a genuine setup issue.
What is the best MMP for a small app team just getting started with attribution?
Adjust, AppsFlyer, and Branch are the three most widely used MMPs and all are viable starting points, but the right choice depends on your budget, tech stack, and ad network mix. AppsFlyer is often recommended for teams running campaigns across multiple ad networks due to its broad integration library, while Adjust is a strong choice for teams that prioritise clean UX and straightforward reporting. Most MMPs offer a free tier or trial, so it is worth testing one against your existing setup before committing — and making sure it supports SKAN configuration if you are running iOS campaigns.
Can I use Google Analytics or Firebase instead of a dedicated MMP?
Firebase (Google Analytics for Firebase) can handle basic app install tracking and works well for teams running exclusively on Google Ads, but it has significant limitations for cross-channel attribution. It does not provide the same level of multi-platform deduplication, SKAN postback management, or granular lookback window control that a dedicated MMP offers. If you are running paid campaigns across more than one platform, a dedicated MMP will give you far more reliable data and save you significant time reconciling numbers manually.
What is a SKAN conversion value schema and how should I set one up?
A SKAN conversion value schema is a mapping that tells Apple’s SKAdNetwork what user actions to encode into the limited conversion signal it sends back to ad networks after an iOS install. Because SKAN only allows a small numeric value with no user-level detail, you need to decide upfront which in-app events or revenue thresholds are most valuable to measure — for example, completing onboarding, making a first purchase, or reaching a specific engagement milestone. Setting this up correctly means prioritising your highest-signal events within SKAN’s constraints, and it should be revisited whenever your key user journey or monetisation model changes. Your MMP’s documentation will have schema-building tools, but aligning it with your actual business goals requires a clear understanding of your user funnel.
Why do my re-engagement campaigns make my install numbers look inflated?
Re-engagement campaigns target users who have already installed your app, but some ad platforms count a re-opened or re-activated session as a new conversion if the user falls within their attribution window. Your MMP, however, typically identifies the device as an existing user and classifies the event as a re-engagement rather than a new install, which removes it from your new install count. This creates a visible gap between platform-reported installs and MMP-reported installs, especially if re-engagement is a significant part of your media mix. The fix is to ensure your MMP has re-engagement campaigns correctly tagged and that you are comparing platform data against the right MMP report segment — new installs versus re-attributions.
How often should I audit my app attribution setup?
A full attribution audit is worth running at least once per quarter, and also any time you launch a new ad network, update your app’s SDK, change your attribution windows, or notice a sudden shift in your discrepancy levels. Smaller checks — such as verifying that key in-app events are firing correctly and that time zones are aligned — should be part of your regular campaign review process, ideally monthly. Attribution setups drift over time as platforms update their APIs and your app evolves, so treating it as a one-time configuration rather than an ongoing discipline is one of the most common mistakes app marketing teams make.
Is there a way to reconcile iOS and Android performance data when iOS numbers are so incomplete?
Direct one-to-one comparison between iOS and Android performance is unreliable given SKAN’s data limitations, but there are practical ways to work with the gap. Use your Android data as a benchmark for understanding funnel conversion rates and payback periods, then apply those benchmarks directionally to your iOS modeled data rather than treating iOS numbers as precise. Some MMPs also offer probabilistic modelling and aggregated measurement solutions that partially fill the SKAN gap. Where possible, focus your iOS optimisation on downstream signals like revenue and retention rather than install volume alone, as these are more stable indicators of true campaign quality under privacy constraints.
Related Articles
- Why did my app installs drop suddenly?
- 6 ways to tell your app attribution setup is broken
- 5 essential steps to advertise your app before launch day
- Can you advertise an app without using Google Ads?
- How do you advertise an app using push notifications and re-engagement?
- How do you advertise an app to reduce user churn?
- What is the difference between organic and paid app installs?
- How do you optimize an app title for the app store?
- What is the difference between Apple App Store and Google Play Store optimization?
- What does ASO stand for in online marketing?
EN 
NL 