Case #019: The conversions I gave away

Case #019: The conversions I gave away

An ecommerce offer, $3,000 budget, $38 average order, and a frustrating start. Days 1 through 6 logged $1,400 spent and $1,090 back. The product was good, the creative tested well in comments, but the tracked ROI sat stubbornly negative and I couldn't find the leak.

The leak was a setting. The offer ran a 1-day click attribution window — if a user clicked my ad and bought 26 hours later, that sale belonged to nobody, or worse, to whatever channel touched them last. For a $38 considered purchase, one day is nothing. People click, think, come back tomorrow, and buy. My window was closing the door on my own converters.

I couldn't change the advertiser's payout terms, but I could prove the gap. I dropped a 7-day cookie of my own on the landing page and matched purchases back to clicks. The 1-day window was capturing only 58% of sales my traffic actually drove. Forty-two percent of my conversions were happening in the 1-to-7 day shadow and getting credited elsewhere.

With that data I renegotiated the offer to a 7-day window. Same traffic, same spend — the tracked revenue caught up to the real revenue.

Full arc after the window change, 15 days total: $3,000 spent, $4,650 back. 55% ROI — most of it conversions I was already generating and quietly donating to other channels.

The lesson: a short attribution window doesn't lose you sales, it gives them away — for any considered purchase, measure your real shadow before you trust the credited number.