How would you estimate the size of the US pet food market?

A worked bottom-up estimation of the US pet food market with real anchor numbers, a sanity check, and the PM "so what" that separates a hire from a pass.

"Estimate the size of the US pet food market"

This question is a viability probe dressed as a math exercise. The interviewer at Chewy, Amazon, or Instacart is not checking your arithmetic: they are checking whether you can identify a market where real money pools, segment it with structural logic rather than vibes, and then say something useful about what the number means for a product or investment decision. The number is the entry point, not the answer.

Clarify before you calculate

Spend 30 seconds on scope. “By pet food, I’ll mean commercially purchased food and treats for companion animals in the US, excluding prescription vet diets and supplements unless you want me to include them. I’ll estimate annual retail revenue. Does that scope work?” Wait for confirmation. This move costs nothing and shows you know that definitional choices shift the output by 30-40%.

Build the estimate bottom-up

Start from households, not population. The US has roughly 130 million households. About 65% own at least one pet, giving approximately 85 million pet-owning households. (APPA data puts the figure at 95 million; note the assumption aloud and adjust if your interviewer prefers the higher anchor.)

Dog-owning households: 53% of US households own a dog, so about 69 million households. Average annual spend on dog food and treats: roughly $450. Subtotal: 69M x $450 = $31 billion.

Cat-owning households: 39% of US households own a cat, about 51 million. Roughly 15% overlap with dog-owner households, so net unique cat-only households: about 43 million. Average annual spend: roughly $330. Subtotal: 43M x $330 = $14 billion.

All other pets (birds, fish, small mammals, reptiles): the remaining ~10 million pet-owning households, averaging about $100/year in food spend. Subtotal: $1 billion.

Running total: ~$46 billion.

Sanity check

APPA reported $68.3 billion in pet food and treat sales for 2024, with a 2026 projection around $69.7 billion. The gap between your $46B and $68B is a definitional difference, not a logic error. Treats alone account for roughly 20-25% of the food budget when counted as a separate line item; adding that to the $46B estimate lands closer to $56B. Mordor Intelligence cites $77-82B for 2025-2026 because their definition absorbs supplements and functional nutrition products. Say all of this explicitly: it shows the interviewer you understand measurement variance and are not confused by your own number.

The PM move: so what?

A $50-70B market with fresh/raw pet food growing at 21% CAGR through 2030 tells you where margin lives. The question for a product team at Chewy or Amazon is not total market size; it is whether you can capture share in the premium segment where LTV is higher and churn is lower once a brand earns trust.

The JTBD lens sharpens the segmentation. Dog owners are not buying food; they are hiring food for three distinct jobs: convenience, health proxy (“this food extends my dog’s life”), and emotional bond (“feeding my pet is a ritual of care”). The health-proxy and emotional-bond jobs are where the 21% CAGR growth lives, and where AI-personalized nutrition by breed, age, and health profile is genuinely viable. Spend is already high, emotional stakes are real, and subscription retention is durable once trust is earned. The sizing exercise becomes a product opportunity brief: not “the market is $60B,” but “there is a fast-growing premium wedge where the viable product earns the health-proxy job.”

Structure a strong answer

strong

"Before I start, one clarifying question: by pet food should I include treats, or food only? And US retail revenue, not wholesale? [Nod.] Great. I'll build this bottom-up from households. The US has roughly 130 million households. About 65% own at least one pet, so roughly 85 million pet-owning households. I'll note the APPA figure is closer to 95 million, but I'll use 85 as a conservative anchor and note the range.

Dog owners: 53% of US households, so about 69 million households. Average annual spend on food and treats, around $450. That gives me roughly $31 billion. Cat owners: 39% of households, about 51 million, but with a 15% overlap with dog households, I'll net that to 43 million unique cat-primary households at $330/year: $14 billion. Other pets are a smaller tail, about 10 million households at $100/year: $1 billion. Total: $46 billion food-focused, or about $56 billion if I add a treat line as a separate 20% premium. APPA's $68 billion figure includes supplements and a broader treats definition, so I'm inside the reasonable range.

The 'so what' for a product team: the fresh and raw segment is growing at 21% CAGR. The premium end is where LTV is higher and churn is lower once trust is earned. If I were sizing next, I'd go one level deeper into that segment specifically, because that is where the viable product opportunity is, not in the commodity dry-food shelf."

weak

"The US has 330 million people. Maybe each person spends about $150 a year on pet food. So the market is around $50 billion." This skips pet ownership rates entirely, produces a nonsensical per-capita number for the 35% of households with no pets, uses a single blended spend figure that ignores the 35% difference between dog and cat spend, and stops at the number with no sanity check and no reflection on what it means. Interviewers read this as someone doing arithmetic theater, not structured estimation. Stopping at the number is the most common failure: "So the market is about $50 billion" with a full stop reads as someone who thinks this is a math test rather than a product judgment exercise.

Where candidates fail

Skipping pet type segmentation is the most common structural error: dog and cat spend differ by about 35%, and blending them flattens the one distinction that shows the interviewer you segment by logic rather than guesswork. Ignoring the $46B-$82B source variance makes you look confused when an interviewer cites a different figure. And stopping at the number, “the market is about $50 billion,” with no reflection on what to do with it, reads as an analyst answer rather than a PM one.

For more on market sizing structure, see TAM/SAM/SOM and the related estimation question on EV market size.