866 euros, one sale. The mandate I would not take again.

I keep this file because I think it is the most useful failure I have, and because the numbers in it are clean.

The numbers, as they are

23,592 impressions. 176 landing page views. 9 add-to-carts. 3 checkouts started. 1 sale. A return on ad spend of zero, against a target of 2.

The funnel lost 99.25 % of its volume between the impression and the page view. The problem was not the checkout. It was upstream of the shop.

What I would still defend

I questioned the brief I had been given. Six weeks in, I wrote a repositioning note that opened on my own mistake: we were producing means before knowing the strategy. In it I named the two levers the brand did not control. The price, set by the supplier. And distribution, imposed by the dropshipping model.

I built the media model on observed rates, not comfortable ones. It showed a break-even click-through rate of 6.75 % against 0.93 % observed, a factor of 7.3, and a gap between spend and revenue that stayed negative every month from October to April. I handed it over with its own proof that it did not close.

I kept the reports that embarrassed everyone. Two reports from the acquisition agency announced spectacular returns. They were the tool's demonstration data: the ad account connection was broken and the dashboard fell back on its examples. I archived them with that explanation.

What I got wrong

I accepted a marketing leadership role on a quarter of marketing. Price and distribution were out of my hands from the first day. I knew it, I wrote it, and I stayed.

I diagnosed after producing. Content, site and influencers left before the target was settled. My June note denounces exactly the error I had made in May.

I read a month of reports before checking the instrumentation. A watermark on the first page was enough to see it.

The target changed without a written decision. A man in the first content calendar, a woman over 45 in the second. Nobody settled it, me included.

The rule I kept

Since then, I only take mandates where I hold the levers that decide the result: the price, the distribution, the budget. When I cannot hold them, I say so before the start, and I propose a different scope instead of a bigger campaign.

That rule is now the first of the five clauses I write into a contract.