Chewy Says 83% of Its Sales Are 'Autoship.' Read That Number Again.
Chewy's flywheel looks unstoppable: 83.3% of its $12.6B in FY2025 sales come from Autoship. But the metric counts every dollar an enrolled customer spends - not just the scheduled refills. The true subscription share is undisclosed, and that gap is the whole story.
Comes with a free Flywheel Designer Canvas template — plus a worked example for Chewy.
A bag of dog food shows up on the same Tuesday it always does, before the old bag runs out, with a hand-written card and sometimes a portrait of the dog the owner never asked for. The customer doesn't shop. The customer doesn't decide. The customer barely remembers a transaction happened. Multiply that quiet refill across 21.3 million active customers6 and you have the engine Chewy points to with the most-quoted number in its filings: 83.3% of fiscal 2025's $12.6 billion in net sales came from Autoship.6 It is, on its face, one of the most beautiful flywheels in retail. It is also not quite what the headline says it is.
The official story is that Chewy is an 83%-subscription business - a recurring-revenue machine wrapped in a pet store. Read Chewy's own SEC definition and that story bends. 'Autoship customer sales' does not mean scheduled subscription orders. It means every dollar spent by anyone who happens to have an active Autoship subscription - including the impulse chew toy, the one-off bag of cat litter, the spontaneous order that has nothing to do with the auto-replenishment schedule.5 The number isn't fake. It's just measuring a wider thing than the word 'Autoship' makes you picture.
“Autoship customer sales [are] sales from Autoship program purchases AND purchases made outside the Autoship program by Autoship customers.”5
The flywheel is real - and the metric is wider than the wheel
Here is the thesis, plainly: Chewy's Autoship flywheel is genuinely turning, but the 83.3% figure flatters it by counting the rider's whole basket, not just the wheel itself. The true subscription-only share - the purely mechanical, set-it-and-forget-it portion - is undisclosed and is necessarily lower. That gap matters, because the entire investment case for Chewy rests on how much of that 83% is locked-in, predictable, and hard to walk away from, versus how much is ordinary e-commerce that happens to belong to a customer who once clicked 'subscribe.' The first kind is a moat. The second kind is just retail with a nicer cohort.
Watch the number climb and you can see the definition doing some of the work. At IPO filing in 2019, roughly 66% of revenue came from Autoship subscription customers.8 By Q1 FY2023 it was nearly 75%.5 By Q3 FY2024, 80%.8 By FY2025, 83.3%.6 A rising share of a customer's total spend being attributed to Autoship can mean two very different things - more people genuinely on recurring schedules, or the same enrolled customers simply buying more of everything else from Chewy. Both are good for Chewy. Only one of them is the durable, churn-resistant subscription moat the headline implies.
| Scheduled auto-replenishment | Same customer's other orders | A non-enrolled customer's orders | |
|---|---|---|---|
| Counted in the 83.3% figure | Yes | Yes | No |
| Genuinely recurring / locked in | Yes | No | — |
| Disclosed separately | No | No | — |
| Drives the 'subscription moat' narrative | Should | Gets credited anyway | — |
On $12.60 billion of FY2025 net sales, Chewy attributes $10,497.1 million to Autoship customers.6 But the numerator bundles two things that behave very differently: predictable refills you'd struggle to cancel, and ordinary purchases that merely belong to enrolled accounts. Chewy never splits them. So the published 83.3% is an upper bound on the part that is actually a subscription - and the gap between the headline and the truth is exactly the part an analyst can't see.
Why the service obsession is the actual flywheel
Strip the metric debate away and something undeniable remains: Chewy turned pet supplies - a commodity Amazon could ship just as fast - into a relationship people don't switch out of. That is the loop. Reliable refills lower the friction of staying; the lowered friction raises lifetime value; the higher value funds the famous service theatrics - the condolence flowers when a pet dies, the surprise pet portraits, the 24/7 humans on the phone - which deepen the emotional cost of leaving; and that stickiness makes the next refill even safer to schedule. Notice what does the compounding here. It isn't price. The Autoship discount itself is funded by vendors, not by Chewy's own margin.8 Chewy isn't buying loyalty at the cost of its P&L; it's renting shelf space in the customer's habit and letting the supplier subsidize the coupon. That is a far better flywheel than a discount war.
Isn't 83% still an enormous moat, however you count it?
The fair objection is that the definitional quibble is a distinction without a difference. Even if a chunk of that 83.3% is discretionary spend by enrolled customers, those customers are still Chewy's customers, buying more, more often - and a wallet-share that high is a moat by any honest reading. There's truth in that. Net income of $222.8 million and $719 million of adjusted EBITDA on $12.6 billion of sales6 are not the numbers of a fragile business. But the steelman cuts both ways. The reason the gap matters is fragility, not size. Recurring subscription revenue is valued at a premium precisely because it's hard to churn - it survives recessions, distraction, and a slightly cheaper competitor. Discretionary spend by a loyal customer is softer; it flexes with the economy and with attention. By bundling the two under one growing percentage and never disclosing the split, Chewy lets investors price the whole 83% as if it were the durable kind. The market is paying subscription multiples for a number that is part subscription, part very-good-retail. That's not a lie. It's an ambiguity worth a lot of money.
The most persuasive numbers in a filing are the ones with a custom name - 'Autoship customer sales,' 'engaged users,' 'community-adjusted EBITDA.' A custom metric is a choice, and the choice usually flatters. The discipline is simple: every time a company leads with a proprietary figure, find the sentence where it defines the term, then ask what the metric is quietly bundling that a stricter version would exclude. Chewy's 83.3% isn't deceptive - the definition is right there in the filing. But the headline travels far and the footnote doesn't. The gap between what the word makes you picture and what the metric actually counts is exactly where the real analysis lives - and where the mispricing hides.
Chewy built something genuinely rare: a commodity business people don't leave, run on a refill nobody thinks about and a service culture nobody expected. The flywheel is real, and it spins. But the 83.3% it points to is a wide-angle lens on a narrower wheel - it photographs the loyal customer's entire basket and labels it 'subscription.' The durable core is in there. Chewy just never tells you how big it is. And in a business valued on the assumption that the answer is 'almost all of it,' the one number Chewy doesn't disclose may be worth more than every number it does.
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 was co-founded by Ryan Cohen and Michael Day in June 2011, originally under the name 'Mr. Chewy,' in Dania Beach, Florida.Wikipedia, Chewy (company) ↗ · 2026
- 2PetSmart acquired Chewy for $3.35 billion in April 2017, which was the largest e-commerce acquisition in history at that time, exceeding Walmart's $3.3 billion acquisition of Jet.com in 2016. Chewy reported nearly $900 million in revenue in 2016 (its fifth year of operation).
- 3Chewy's S-1 prospectus (filed April 29, 2019) shows net sales of $900,566K in fiscal 2016, $2,104,287K in fiscal 2017, and $3,532,837K in fiscal 2018, confirming the revenue trajectory at time of IPO filing.
- 4Chewy completed its IPO on June 18, 2019. The IPO dual-class structure — Class B shares carrying ten votes each versus one vote for Class A — gave PetSmart approximately 77% of the combined voting power post-IPO.[[cite:s9]]
- 5In Q1 Fiscal 2023 (ended April 30, 2023), Autoship customer sales represented nearly 75% of total net sales, with net sales of $2.78 billion up ~15% year over year. Chewy defines Autoship customer sales as sales from Autoship program purchases AND purchases made outside the Autoship program by Autoship customers — a critical definitional nuance.
- 6For fiscal year 2025 (ended February 1, 2026), Chewy reported net sales of $12.60 billion (up 6.2% YoY, or 8.3% on a normalized 52-week basis), Autoship customer sales of $10,497.1 million (83.3% of net sales), 21.327 million active customers (up 4.0% YoY), net sales per active customer of $591, net income of $222.8 million, and adjusted EBITDA of $719 million.
- 7Chewy's first outside investment was $15 million from Volition Capital in 2013. Ryan Cohen says he was rejected by over 100 venture capital firms before that investment (attributed-to-source; not independently verified).Wikipedia, Ryan Cohen ↗ · 2026
- 8At IPO filing time (2019), approximately 66% of Chewy's revenue came from Autoship subscription customers. By Q3 FY2024, Autoship customer sales represented 80% of net sales ($2.3 billion, growing 8.7% YoY). The Autoship program offers customers a discount funded by vendors, not by Chewy directly.
- 9Following the IPO, PetSmart held approximately 77% of the combined voting power of both classes of Chewy common stock outstanding immediately after the offering.
- 10In Q3 FY2024, Autoship customer sales reached $2.3 billion, an 8.7% increase year-over-year, representing 80% of net sales for the quarter.