Case study: fixing duplicate postbacks recovered 9% phantom conversions
What it is: A forex media buyer reconciled tracker numbers against broker payout reports and found a 9% gap. The cause was duplicate FTD postbacks firing on both deposit and KYC-approval events.
Best for: Anyone running self-hosted Keitaro or RedTrack against multiple broker integrations.
What was done:
— Pulled 3 months of postback logs (about 4,100 conversion events)
— Matched click_id to broker transaction_id, not to email
— Added a dedup rule keyed on transaction_id with a 30-day window
Outcome:
— Removed 372 phantom conversions that had inflated reported FTDs
— Reported ROI dropped 9% on paper, but the real number was finally trustworthy
— Two campaigns that looked profitable were actually at -14% and got cut
Pros:
— transaction_id keying survives users who re-register with a new email
— One dedup rule fixed every broker feed at once
Cons:
— Keitaro's default dedup is click-based, so this needed a custom S2S rule
— Brokers that don't send transaction_id forced a fallback to email+timestamp
Who should skip this: Single-offer buyers with one clean integration — your duplicate risk is near zero.
Verdict: Trust transaction_id, not email.
Spread Bench
@SpreadBench
Case study: fixing duplicate postbacks recovered 9% phantom conversions
Этот пост опубликован в Telegram-канале Spread Bench. Подписаться можно по ссылке: @SpreadBench.