Chewy's Famous 83% Isn't a Subscription Number. Read the Footnote.
Chewy says 83.3% of its $12.6 billion in sales now comes from 'Autoship customers.' It's the most-cited stat in the pet-retail world - and it counts a customer who tried Autoship once, cancelled, and bought nothing on schedule for a year.
Comes with a free Flywheel Designer Canvas template.
Chewy has one number it puts in front of every investor, every quarter, in bold: in fiscal 2025, 83.3% of its $12.60 billion in net sales came from Autoship customers.3 It is the most-cited statistic in pet retail, and it is meant to sound like a confession of strength - that more than four out of five dollars now arrive on a recurring, predictable, set-it-and-forget-it schedule. Bags of kibble landing on doorsteps like clockwork, no human decision required. It is a beautiful number. It is also not measuring what almost everyone thinks it is measuring.
The story the number tells is 'Chewy is becoming a subscription business.' The story buried in the footnote is something quieter and stranger - and it is the company's own definition, filed with the SEC, that gives it away.
“Autoship customer sales are all sales — both through and outside the Autoship program — from customers who had at least one Autoship order shipped in the preceding 364-day period.”1
The 364-day trick hiding in plain sight
Read that definition slowly, because everything turns on it. The metric does not count scheduled recurring deliveries. It counts every dollar a person spends once they have triggered a single Autoship order at any point in the prior 364 days.1 Picture the customer Chewy dangles a 35% first-order discount in front of - capped at $20 in savings.6 They take the deal, get their cheap first bag, and cancel the next morning. No fee, no penalty, no minimum; Chewy's own terms make cancellation a one-click affair.7 Then for the next year they buy from Chewy the old-fashioned way: one-time orders, whenever the dog runs low. Every one of those one-time purchases is booked as 'Autoship customer sales.' The recurring engine never ran. The number says it did.
So the headline figure is not a subscription-revenue figure at all. It is a loyalty-cohort figure dressed in subscription clothing - the spending of everyone who ever touched Autoship, recurring or not. That distinction is the whole game, because subscription revenue and 'revenue from people who once subscribed' are governed by completely different forces. One is contractual momentum. The other is just brand preference, which any competitor can attack.
| What people assume | What Chewy actually counts | |
|---|---|---|
| The unit | Scheduled recurring deliveries | All sales by anyone who ordered Autoship once in 364 days |
| The trial-and-cancel customer | Excluded after they cancel | Fully counted for a year |
| One-time orders by enrolled buyers | Excluded | Counted as Autoship sales |
| What it proves | Locked-in recurring revenue | Brand preference within a loyal cohort |
The flywheel is real — that's why the number can afford to be loose
Here is the part the skeptics miss: the underlying machine genuinely works. Autoship-customer concentration has climbed every year - 76.2% of net sales in FY2023, 79.2% in FY2024, 83.3% in FY202523 - and that trend line is not an accounting artifact. The mechanism is straightforward and durable. A 35% first-order hook gets a pet owner into the program cheaply, a portion of which is funded not by Chewy's own margin but by vendors, who pass discount dollars through to soften the economics.56 Once the kibble arrives on a schedule, the customer stops shopping - the cognitive cost of comparing prices on a recurring necessity is higher than the few dollars they might save - and Chewy quietly becomes the default. The flywheel runs on inertia, and inertia is cheap to maintain.
And the demand is structural, not Chewy-specific: roughly 49% of online pet-product buyers used an Autoship-style subscription in the prior twelve months, per a survey Chewy itself cites.4 Pet food is the perfect subscription category - bought relentlessly, predictably, by people who feel guilt about running out. Chewy did not invent that demand. It built the smoothest pipe to capture it. So the flywheel is real, the concentration is real, and the strategic logic is sound. The problem is purely one of measurement - and measurement is where management gets to choose what you see.
What's missing tells you more than what's shown
The cleanest signal in any disclosure is the number a company could publish and doesn't. Chewy reports neither a churn rate nor the share of 'Autoship customer sales' that actually flows through scheduled recurring orders versus one-time purchases by enrolled buyers.8 If 80% of that $10.50 billion truly arrived on a recurring schedule, that figure would be on the first slide - it is the single most flattering thing a retailer could say. Its absence is not proof of weakness, but it is a tell. You report the metric that looks best, and you define it to look as good as it can. The 364-day window, the inclusion of one-time orders, the silence on churn: each is a defensible choice that all happens to push the same direction.
Look closer and the seam shows. In FY2023, Autoship penetration rose while active customers actually fell - from 20.4 million to 20.1 million, a -1.6% decline year over year.910 A pure recurring-revenue flywheel that was genuinely locking customers in should not lose customers at all. What you see instead is a cohort metric rising as the base shrank slightly: the loyal core spent more of its wallet at Chewy even as the edges churned away. That is a real and valuable dynamic. It is just not the unbreakable subscription lock the headline implies.
When a company leans on one proprietary metric - 'Autoship customer sales,' 'monthly active users,' 'annual recurring revenue' - go straight to the footnote where it defines the term, not the slide where it celebrates it. The gap between the intuitive meaning and the actual definition is where management does its quietest work. Two questions cut through almost every constructed metric: What does this include that I'd assume it excludes? And what obvious, more honest number is conspicuously not here? Chewy's flywheel is genuine - but the figure measuring it is built to flatter, and the churn rate that would settle the question stays in the vault. A metric you can't audit is a story, not a fact.
None of this makes Chewy a weak business. Its retention engine is real, its category is sticky, and 83.3% concentration in a loyal cohort is a genuine asset - the honest objection is that even a loosely defined metric can describe a strong company, and it does. But strength and the measurement of strength are two different things, and Chewy has quietly conflated them. The flywheel spins. The question its own disclosure refuses to answer is how much of it is momentum, and how much is just the same loyal owners reaching for the same bag of food - which they'd buy with or without a subscription. Chewy knows the answer. That's exactly why it doesn't print it.
Flywheel Designer Canvas
A one-page canvas for mapping a business's flywheel: the reinforcing loop, how it was started, the second-order loops it spins off, the moat it creates, and how it could spin backward. Use it to diagnose whether you have a real flywheel or a funnel drawn in a circle — and to design one of your own.
The worked example unlocks with a subscription. See plans →
Sources
Where this comes from — the filings, records, and reporting behind it.
- 1Chewy defines 'Autoship customer sales' as all sales—both through and outside the Autoship program—from customers who had any Autoship order shipped in the preceding 364-day period; it is NOT a pure scheduled-delivery revenue figure.
- 2In fiscal year 2024 (ended Feb. 2, 2025), Chewy reported $9.39 billion in Autoship customer sales, representing 79.2% of $11.86 billion in total net sales, up from 76.2% in FY2023.
- 3In fiscal year 2025 (ended Feb. 1, 2026), Chewy reported $10.50 billion in Autoship customer sales, equal to 83.3% of $12.60 billion net sales, up from 79.2% in FY2024; active customers rose 4.0% to 21.327 million.
- 4Chewy's FY2025 Form 10-K (for the year ended Feb. 1, 2026) is on file with the SEC and corroborates Autoship customer sales of $10.50 billion at 83.3% of net sales; the 10-K also discloses that approximately 49% of online pet product buyers used an Autoship/subscription program in the prior 12 months per a January 2025 Packaged Facts survey.
- 5Chewy's FY2024 10-K explicitly discloses that 'certain vendors provide funding for discounts relating to the Autoship subscription program which are passed on to the Company's customers'—meaning Autoship discounts are partly supplier-subsidized.
- 6The first Autoship order carries a 35% discount (capped at $20 in savings); subsequent recurring orders on select brands receive 5% off. There is no membership fee for Autoship itself.
- 7Chewy's Autoship Terms confirm customers can cancel at any time via Autoship Manager or by phone with no cancellation fees or penalties; Chewy reserves the right to change discount amounts at any time.
- 8Motley Fool analysis (Feb. 2026) adversarially argues that 'Autoship customer sales' is a 'murky metric': it counts all purchases by anyone who enrolled in Autoship in the prior 364 days, including non-Autoship orders and potential trial-and-cancel customers; Chewy does not separately disclose actual scheduled-delivery revenue or churn rates.
- 9At the end of FY2022 (January 29, 2023), Chewy reported 20,405 thousand active customers — the prior-year baseline against which FY2023's 20.1M (-1.6% YoY) decline is measured.
- 10At the end of FY2023 (January 28, 2024), Chewy reported approximately 20,082 thousand active customers — the year-over-year decline from FY2022's 20,405 thousand baseline.
- 11In fiscal year 2024 (ended February 2, 2025), Chewy reported approximately 20.5 million active customers at year-end, consistent with FY2025's reported 4.0% YoY growth to 21.327 million.