Playbook: Reconcile MMP Numbers vs Self-Attributing Networks

Playbook: Reconcile MMP Numbers vs Self-Attributing Networks

Google, Meta, TikTok grade their own homework. Reconcile them properly:
— Accept the model: SANs report their own attributed installs; your MMP arbitrates last-touch
— Pull both numbers for the same window and expect a gap — SANs over-claim by design
— Set a consistent attribution window across MMP and each SAN, or the gap is just window mismatch
— De-duplicate cross-network: one install can be claimed by two SANs; your MMP picks one winner
— Compare cost figures from the network with installs from the MMP — never mix sources for CPI
— Document the typical variance per network so a 30% SAN over-count doesn't trigger a false alarm
— Use MMP-attributed installs as your single source of truth for budget decisions

✓ Stops you double-paying for double-claimed installs
✓ Establishes one arbiter instead of three conflicting dashboards
✗ SANs limit MMP visibility, so reconciliation is never perfect
✗ Window and de-dupe rules differ subtly per network

Verdict: Use the MMP as arbiter for cross-network decisions; treat SAN self-reported installs as upper-bound marketing math.
Best for: Buyers splitting budget across self-attributing in-app networks.