Why Facebook and TikTok Report Different Install Numbers

You run campaigns on both Facebook and TikTok.

Facebook reports 500 installs. TikTok reports 300 installs. Your MMP shows 650 total installs across both platforms.

The math doesn't add up. And it never will.

Here's why Facebook and TikTok will always show different numbers—and why comparing them directly doesn't make sense.

The Core Problem: Self-Attribution

Both Facebook and TikTok self-attribute installs.

What this means:

If a user clicks a Facebook ad within the lookback window and installs your app, Facebook claims that install—regardless of what else the user clicked.

If the same user also clicked a TikTok ad within TikTok's lookback window, TikTok also claims that install.

Both platforms count it. Your MMP only counts it once, giving credit to whichever click was last.

Example scenario:

Day 1: User clicks Facebook ad Day 2: User clicks TikTok ad Day 3: User installs

Facebook's view: Install attributed to Facebook (click within 7-day window)

TikTok's view: Install attributed to TikTok (click within 7-day window)

Your MMP's view: Install attributed to TikTok (last click wins)

Facebook reports it. TikTok reports it. Your MMP counts it once for TikTok.

This is the primary reason platform numbers don't add up.

Reason 1: Different Attribution Logic

Facebook:

• Self-attributes if user clicked or viewed an ad within the attribution window
• Includes view-through attribution (user saw ad but didn't click)
• Uses internal attribution model optimized for Facebook's ecosystem

TikTok:

• Self-attributes if user engaged with an ad within the attribution window
• Has its own view-through logic
• Optimizes for TikTok's behavioral signals

Your MMP:

• Uses last-click attribution across all channels
• Deduplicates overlapping claims
• Provides single source of truth

When multiple platforms claim the same install, they all report it. Your MMP arbitrates and assigns it to one source.

Reason 2: Reporting Timestamp Differences

Platforms report conversions using different timestamps.

TikTok:

• Reports conversions by the timestamp of the ad interaction (click or impression)
• Install on Day 5 from a Day 3 click gets reported on Day 3
• Numbers shift as installs from old clicks arrive

Facebook:

• Also reports primarily by interaction timestamp
• Has separate views for "actions by click date" vs "actions by conversion date"
• Default dashboard shows click date, creating rolling discrepancies

Your MMP:

• Reports by install date
• Install on Day 5 gets reported on Day 5, regardless of when click happened

This timing difference creates daily mismatches that balance out over longer periods.

Reason 3: Lookback Window Variations

Each platform has default attribution windows that may not match your MMP configuration.

Typical defaults:

• Facebook: 7-day click, 1-day view
• TikTok: 7-day click, 1-day view
• Your MMP: Configurable, often 7-day click, 1-day view

If your MMP uses a 7-day click window but you accidentally set a 28-day click window on Facebook, Facebook will claim installs from older clicks that your MMP doesn't attribute.

Common mismatch scenario:

User clicks ad on Day 1, installs on Day 9.

With 7-day window: Not attributed (click too old)

With 28-day window: Attributed to the click

Misaligned windows create systematic over- or under-reporting on specific platforms.

Reason 4: Data Sharing and Postback Timing

TikTok and Facebook receive attribution data differently.

Facebook:

• Receives postbacks from your MMP when installs are attributed to Facebook
• May process and display data with slight delays
• Has mature MMP integration infrastructure

TikTok:

• Migrated to Self-Attributing Network (SAN) integration in 2025
• Legacy MMP integrations deprecated March 31, 2025
• Transition period created temporary data discrepancies

During migration windows or integration updates, temporary mismatches between what platforms report and what your MMP sees are common.

Reason 5: Conversion Event Definition

Platforms may count events differently.

"Install" can mean:

• App download (what the App Store counts)
• First app open (what most MMPs count)
• Install + SDK initialization (what some platforms count)
• Attributed install (what Facebook counts)

If there's a delay between download and first open, or if users download but never open, numbers diverge.

Reason 6: SAN Integration Changes (2025)

TikTok's move to Self-Attributing Network (SAN) integration changed how data flows.

Previous model:

• MMP sent postbacks to TikTok for attributed installs
• TikTok reported based on MMP data

Current SAN model:

• TikTok self-attributes and shares data with MMP
• MMP deduplicates TikTok's attributions against other sources
• Creates new opportunities for discrepancies

Early in this transition (2024-2025), many advertisers saw larger-than-normal discrepancies as systems adjusted.

Reason 7: Fraud Filtering Differences

Your MMP applies fraud detection. Ad platforms may or may not filter the same traffic.

MMP fraud filters block:

• Click flooding
• Install farms
• Emulator traffic
• Suspicious IP patterns

Facebook and TikTok have their own fraud systems, but they don't align perfectly with MMP filters.

Result:

TikTok reports 100 installs. Your MMP receives 100 install postbacks but flags 10 as fraud. Your MMP shows 90 TikTok installs.

TikTok still reports 100 because they weren't filtered on their side.

What's Normal

Expected discrepancy range: 5-15%

Small variances are inherent to the system. Factors:

• Self-attribution overlap
• Timing differences (usually balances out weekly)
• Minor fraud filtering differences
• Postback delivery delays

Problem threshold: 20%+ or systematic direction

Large discrepancies indicate:

• Misaligned attribution windows
• Postback configuration errors
• SDK implementation issues
• Significant fraud

How to Diagnose Discrepancies

When Facebook and TikTok numbers diverge significantly from your MMP:

1. Check attribution window alignment

• Log into Facebook Ads Manager: Settings → Attribution
• Check TikTok Ads Manager: Settings → MMP Integration
• Verify your MMP has matching windows configured for each platform

2. Compare over longer periods

• Daily numbers fluctuate due to timing differences
• Compare 7-day or 30-day totals for more stable comparison
• Account for SKAN delays (3+ days for iOS data)

3. Verify postback configuration

• Check MMP dashboard for postback success rates
• Look for failed postbacks or error rates
• Test postback URLs using platform validation tools

4. Review event definitions

• Confirm all systems define "install" the same way
• Check if platforms count app download vs first open differently
• Verify SDK fires on first open, not just download

5. Check for integration updates

• TikTok SAN migration (2025) required reconfiguration
• Facebook SDK updates sometimes require MMP SDK updates
• Review release notes for both platforms

Which Number to Use for What

Different platforms serve different purposes:

Use Facebook's numbers for:

• In-platform optimization (Facebook's algorithm uses its own data)
• Creative testing within Facebook
• Budget pacing within Facebook campaigns

Use TikTok's numbers for:

• In-platform optimization
• TikTok-specific creative performance
• Budget pacing within TikTok

Use your MMP's numbers for:

• Cross-channel budget allocation
• True ROAS and LTV analysis
• Attribution comparison across all channels
• Financial reporting and decision-making

Never compare Facebook's reported installs directly to TikTok's. Compare both to your MMP's attributed counts for each platform.

Reducing Discrepancies

You can minimize but never eliminate discrepancies:

1. Align attribution windows

Set identical lookback windows in Facebook, TikTok, and your MMP.

2. Use consistent event definitions

Ensure "install" means the same thing across all systems.

3. Implement SDKs correctly

Follow MMP integration guides precisely. Test in development environments.

4. Monitor postback health

Check your MMP dashboard weekly for postback delivery rates and errors.

5. Update integrations

When platforms change (like TikTok's SAN migration), update configurations immediately.

Even with perfect setup, expect 5-10% variance. Different systems measuring different things will never produce identical numbers.

FAQs

Why do Facebook and TikTok show different install counts?

Facebook and TikTok use different self-attribution models, report by different timestamps, and have different data-sharing agreements with MMPs. Both platforms claim credit for installs independently, leading to different numbers.

Which platform's data should I trust?

Neither. Use your MMP as the source of truth. MMPs deduplicate installs across all platforms using consistent last-click attribution, giving you an accurate view of which channel actually drove each install.

What causes the biggest discrepancies?

The biggest discrepancies come from self-attribution overlap (both platforms claiming the same install), timing differences (click date vs install date reporting), and different lookback windows between platforms and MMPs.

Should Facebook and TikTok numbers add up to my total MMP installs?

No. Because both platforms self-attribute, they'll often claim the same installs. Your MMP deduplicates these, so total MMP installs will be less than Facebook + TikTok combined.

How do I fix large discrepancies?

Check attribution window alignment, verify postback configuration, ensure SDKs are implemented correctly, and confirm you're comparing the same time periods and event definitions across all platforms.

---

Facebook and TikTok will never show matching install counts. That's not a problem—it's how self-attribution works. Use your MMP as the single source of truth for decision-making while using platform data for in-platform optimization.