<b>Null conversion values: the SKAN data you're throwing away</b>
When a source doesn't cross Apple's privacy threshold, the postback arrives with a null (or zero) conversion value. Most teams discard these. That's a mistake.
✓ A null postback still confirms an install happened from that ad network
✓ Counting nulls separately reveals which sources are too fragmented to ever optimize on value
✗ MMPs vary wildly in how they surface vs hide null-value postbacks
✗ Treating null as zero-revenue pollutes your ROAS downward
The tactic: track three buckets per source — non-null with value, null-but-confirmed install, and no postback. Sources stuck in bucket two need consolidation (fewer, bigger placements) to ever cross threshold.
Reviewers' tell: ask any MMP demo to show you null-postback handling explicitly. Many can't, cleanly.
<b>Verdict:</b> count nulls as installs, never as zero revenue, and use them to spot sub-threshold sources.
<b>Best for:</b> iOS analysts fixing distorted SKAN ROAS.
In-App Bench
@InAppBench
<b>Null conversion values: the SKAN data you're throwing away</b>
Этот пост опубликован в Telegram-канале In-App Bench. Подписаться можно по ссылке: @InAppBench.