Church & Dwight

Objective

Move retail media past last click and past the ROAS each platform reports on itself, in order to know which spend actually produces sales that would not have happened otherwise, and to redeploy the existing budget toward those campaigns without asking for more.

The deployment

The starting point is an arbitration problem. On retail media networks, each platform grades its own campaigns and reports a ROAS that also counts sales that would have happened without advertising. Church & Dwight was therefore judging its campaigns on last click, on metrics that arrive late and do not say where to send the next dollar. Three vendors were assembled to answer that question. Incremental produces a daily causal reading at campaign level: the model breaks total sales into three parts, the organic baseline, the effect of digital shelf movements (price, promotions, coupons, ratings and reviews, search position), and what is left, attributed to advertising. That residual share divided by spend gives iROI, incremental return on investment, which Incremental's documentation defines as incremental profit over spend. The method combines econometric modeling, dedicated experiments and quasi-experimental techniques of the difference-in-differences type. Skai brings the other half. A direct API connection between the two platforms pushes Incremental's daily recommendation down into Skai's optimization rules and bidding algorithms, which execute it across thousands of campaigns with no spreadsheet in between. WPP Media, through Wavemaker, holds the strategic guardrails and arbitrates. The most visible operational consequence is the shift from a frozen monthly budget to a weekly reallocation: investment moves each week toward campaigns with high incremental return and pulls back from the rest. Campaign structures were rebuilt to allow granular testing, by SKU, by page type and by product category. The setup was applied to the Batiste, TheraBreath and Arm & Hammer brands, at Target and on Instacart. It won the Grand Prix at The Drum Awards for Commerce Media on April 27, 2026 in Miami, along with the Commerce Media Partnership of the Year award, and Incremental was recognized by the Adweek Tech Stack Awards 2026 in August 2026 on the basis of this case.

Results Proof B

The most precise source is the case study published by Skai on May 22, 2026, which names the brand, the three products, the two retailers and gives the figures with their exact scope. It is an official source, but an interested one: Skai sells the activation platform used in the setup. Incremental, the other vendor, publishes its own figures on the same case. Two trade publications pick up the results, The Drum on April 28, 2026 after its jury, and Adweek on August 4, 2026 in an awards list. None of these texts is independent of the setup it describes: both articles report on award entries. Church & Dwight, for its part, has never confirmed these figures, neither in a press release nor in its financial results. Two discrepancies justify not going higher. First the wording: Skai speaks of media efficiency at Target, while The Drum and Adweek speak of incremental return on investment for the same 292%, with no source saying whether the two measure the same thing. Then the orders of magnitude: Incremental's methodology page claims up to 122% iROI gains across TheraBreath, Batiste and Arm & Hammer Laundry combined, which does not reconcile with the 292% Skai reports on Batiste alone. So the two vendors on the same project do not publish the same figure for the same thing. Church & Dwight's organic growth and group revenue are not counted here: nothing in the sources ties them to this setup.

1. 1
   Rebuild the campaign structures Agency commerce activation team

   Campaigns are re-cut to allow granular testing by SKU, by page type and by product category. Without that split, measurement has nothing to compare.

2. 2
   Daily decomposition of sales Incremental causal model

   The model isolates the organic baseline, then removes the effect of prices, promotions, coupons, reviews and search position. What is left is attributed to advertising.

3. 3
   Compute incremental return Incremental causal model

   The residual share over spend gives iROI per campaign, with response curves showing where spend saturates and where it still has room.

4. 4
   Transmission by API Skai-Incremental connection

   The day's recommendation flows straight into the optimization rules and bidding algorithms of the activation platform, with no manual re-entry.

5. 5
   Automated execution Skai activation platform

   Bids and budgets are changed across thousands of campaigns. The sources state that decisions that used to take weeks are now made and applied within hours.

6. 6
   Weekly reallocation Agency, based on the model's output

   Investment moves each week toward high-return areas and pulls back from low-return ones, with the overall envelope unchanged.

7. 7
   Arbitration and guardrails WPP Media with the brand teams

   The agency sets what the automation is allowed to do, which brands and which retailers fall in scope, and which limits are not crossed.