Criteo Retail Media Allocation Across Retailers

A brand buying eight retailer networks that all run on Criteo receives eight reports that cannot be added together: different click windows, some counting view-through, some crediting the whole basket, some only the promoted item. This retainer restates all of them on one basis, click only, one window, promoted item and its variants, and then makes the monthly allocation call in writing, naming the retailers losing money and by how much. In the last three portfolios I took on, the restatement alone reversed which retailer sat at the top and which sat at the bottom. What the restatement touches: Each retailer's closed-loop report, pulled through the Criteo Retail Media API rather than by hand so the definitions stay visible; the window and view-through settings each retailer applies by default; whether the network credits the promoted item, the brand, or the whole basket; and the share of your volume on that retailer that never passes through the network at all. Retailers that publicly run on Criteo, Carrefour and Best Buy among them, each configure this differently, and those differences are larger than the performance gaps people spend quarters arguing about. How the monthly decision is made: One meeting, one page. Retailer by retailer: contribution per dollar on the restated basis, what changed since last month, what I am moving, and where I am taking it from. Budget moves between retailers rather than between keywords. Keyword-level tuning inside a single retailer is worth a fraction of getting the split between retailers right, and it consumes most of the hours the average shop bills for. Onsite is planned first because it sits closest to the purchase. Offsite extensions of retailer audiences are bought only where they would clear as a standalone prospecting buy at that CPM, which most months they do not. What I will not run: Automated return-target bidding across retailers whose attribution definitions differ, because the algorithm optimises to inconsistent labels and will quietly concentrate your money on whichever network counts most generously. Share of voice or share of shelf as a performance metric. Any commitment given to a retailer's sales team before the numbers have been restated. And I take no percentage of media, so recommending you cut a retailer costs me nothing. Outside the scope: Retailer relationship management and commercial negotiation, which belongs with your sales lead. Creative production to each retailer's specifications. Item content and catalogue feeds. Standing up Commerce Yield if you are on the retailer side, which is a different practice and I will refer it out. And Amazon, which is its own discipline and is not part of this retainer. Who this is wrong for: Portfolios under about thirty thousand dollars a month across all retailers, where the split matters less than my fee. Brands where a single retailer holds more than eighty percent of the volume, who need that one retailer run properly rather than a portfolio decision. And teams where budget cannot actually move between retailers mid-quarter because each pot is committed to a different account manager. That is a commercial problem, and this retainer will only document it. How it runs: Monthly cycle, flat fee, ninety days' notice on either side. I work inside your existing reporting stack rather than building a parallel dashboard for you to maintain after I leave. English or Hebrew.