Pet Food vs Pet Treats, The Divergent Margin Trajectories We Modeled in Q1 2026

340 basis points. That is how much pet food contribution margin compressed across the 9 pet brands we model on Amazon between January and March 2026. Treats, in the same window, expanded 210 basis points. The two sub-categories are diverging fast and the divergence is structural.

If you sell both, and most 7-figure pet brands do, the implication for ad spend, inventory, and new SKU launches in the back half of 2026 is significant. This note walks through what we are seeing, why it is happening, and the one allocation move we have already shifted on three client accounts.

What the Q1 2026 numbers actually look like

Across our pet portfolio, blended food TACoS rose from 14.2% in Q4 2025 to 17.6% in Q1 2026. Blended treat TACoS held flat at 11.1%. Food return rates climbed from 4.8% to 6.1%, driven mostly by formulation-change complaints on three large brands. Treat return rates stayed at 1.9%.

Cost-per-click on food kept rising. Average pet-food CPC on our managed accounts moved from $1.34 to $1.51 quarter over quarter. Treats CPC moved from $0.89 to $0.94, a much smaller climb. Buy box pressure on food intensified as private-label entrants (Amazon Basics Pet, Wag, Wholehearted) added 14 new SKUs across kibble and wet food. Private label pushed into treats too but with fewer new SKUs and weaker review velocity.

Net effect on contribution margin after ads, returns, and FBA fees: food brands lost 340 bps. Treat brands gained 210 bps. The 550-bp swing is the largest intra-pet divergence we have measured in three years of running these accounts.

Why the split is structural, not cyclical

Three forces are driving the divergence and none of them resolve in Q3.

First, food is a frequency category and treats are an emotion category. Food shoppers comparison-shop on price and ingredients. Treats shoppers buy on impulse and brand affinity. Amazon’s algorithm rewards frequency-category competition by giving the buy box to whoever can subsidize the price floor. That is private label or Chewy-affiliated brands, not independent pet brands.

Second, Subscribe & Save penetration is much higher on food (51% of units in our portfolio) than on treats (19%). S&S compresses margin by 5-15% per unit and creates customer-stickiness that the brand never owns. Every food unit shipped on S&S is a unit Amazon is using to lock in its own customer relationship.

Third, Rufus query rewrites favor treat-side flexibility. A shopper asking “best training treat for puppy with sensitive stomach” gets routed to multiple SKUs across multiple price points. A shopper asking “best kibble for senior labrador” gets routed to a much narrower set, and that set is increasingly dominated by either prescription brands or Amazon’s own labels. We covered the Rufus mechanics in our pet supplements 2026 H1 review, same pattern, different SKU class.

The one allocation move we have made on three client accounts

For three pet brands that sell both food and treats, we have shifted approximately 22% of paid budget out of Sponsored Products on food and into Sponsored Brands and Sponsored Display on treats. The thesis: food paid spend is buying flat or declining marginal margin. Treat paid spend is buying expanding marginal margin and higher LTV via brand-affinity capture.

Six weeks in, the data is consistent: the three accounts blended TACoS dropped 1.4 points, treat unit velocity rose 18%, and food unit velocity dropped 3%. Net contribution dollars rose on all three brands. We are not arguing food spend should go to zero. We are arguing the marginal dollar in pet right now is worth more on the treats side, and most brand managers have not adjusted their allocations to reflect Q1’s data.

What this means for new SKU launches in Q3

If you are planning a Q3 2026 launch, the math has shifted. Launching a new kibble or wet food SKU into a 17.6% TACoS category with 6.1% return rates and a Subscribe-and-Save subsidy expectation requires a much larger launch budget than it did in 2024. Treats launches are still working at 11% TACoS with 1.9% returns and a smaller S&S drag.

For brands debating between launching their fifth kibble line or their first soft-chew training treat, the unit economics now strongly favor the treat. We are not saying skip food entirely, house-brand loyalty is still real, but the cost of acquisition on net-new food SKUs has crossed a threshold most brand operators have not yet priced into their launch models.

If you want a 30-minute teardown of your pet brand’s food vs treats P&L on Amazon, email us at hello@clearsightconsulting.com. We will pull your last 90 days, layer in the Q1 benchmarks above, and tell you where the marginal dollar is actually working. No retainer, no deck, one call, real numbers.