Why pet stores overspend when they optimise for ROAS

In pet, the orders that are easiest to win on Google Ads can be the ones that earn the least. Food and litter can make large orders, which is exactly what ROAS rewards. But they are heavy and expensive to ship, and ROAS sees none of that.

The figure above shows how much of the market those orders represent. The American Pet Products Association puts US spending on pet food and treats at $68.3 billion in 2025, out of $158 billion in total pet industry spending. That is more than vet care ($41.0 billion) or supplies, live animals and over-the-counter medicine ($34.4 billion).

How the gap forms

High order value, thin margin

A big bag of premium food makes a large order and a strong ROAS. What the order is worth depends on the margin, and on food that margin has to cover heavy shipping. For context, at Chewy, where consumables made up $8.8 billion of $12.6 billion in net sales in fiscal 2025, cost of goods was 70.2% of net sales.

Shipping cost that ROAS can't see

Google Ads records the order value at checkout. The cost of shipping a heavy order comes out of margin afterwards and never reaches the account.

Repeat orders ROAS doesn't count

The gap also runs the other way. Autoship and subscription reorders mostly arrive without an ad click, so Google Ads doesn't credit them to the campaign that won the customer. Chewy reports that customers on its Autoship programme accounted for 83.3% of its net sales in fiscal 2025. A food campaign that loses money on the first order can still pay back, but only if those customers reorder, and only if you've measured that they do.

Where the gap is smallest

Light hard goods such as toys, collars and bowls cost little to ship relative to their price, so their ROAS and POAS sit much closer together. That makes them easier to judge on a single order, and it's why they need targets of their own rather than sharing one with food.

Working out your own gap

• Split ROAS by product type. Report consumables and hard goods separately, so one doesn't hide the other.
• Put shipping into cost of goods, per order, weighted by what the order actually weighed.
• Track first-order POAS and cohort payback separately. The first tells you what a campaign costs; the second tells you whether its customers stay long enough to cover it.

How to close it

1. Build separate campaigns for consumables and hard goods, with separate targets.
2. Set consumables targets from margin after shipping, not from list margin.
3. Exclude orders that can't pay for their own shipping, such as single heavy items below your free-shipping threshold, unless repeat data says they pay back.
4. Judge food campaigns on payback over time, using your own Autoship or reorder data.