Build-A-Bear teardown: 55% margins, zero debt

Build-A-Bear Workshop is the retail turnaround almost nobody bet on. Written off as a pandemic casualty when malls emptied out, the stuffed-animal brand instead posted five consecutive record-revenue years, expanded gross margin to a level most specialty retailers can only dream about, and did it all while carrying zero long-term debt. When I talk to founders running physical-plus-digital brands, this is the case study I keep reaching for, because the lesson underneath the plush toys is the most important margin question any operator can ask. This teardown works straight from the SEC filings (fiscal years run to late January, so FY2025 ended January 31, 2026) and pulls out what actually applies to a direct-to-consumer (DTC) business.

The mall brand that isn't really a mall brand anymore

Start with the top line, because it's the part that surprises people. Build-A-Bear grew revenue from $411.5M in FY2021 to $529.8M in FY2025. That's five straight record years and a compound annual growth rate of about 6.5%, on a base most analysts assumed had peaked a decade earlier. FY2025 was the first time the company crossed $500M in annual revenue.

The reason it works is that Build-A-Bear runs three engines at once, not one. The first is the workshop itself, a company-operated store where the customer builds a bear as an experience and pays an occasion premium for it. The second is a fast-growing commercial and franchising layer that lets partners fund the stores while Build-A-Bear collects royalties and merchandise revenue. The third is a capital-return machine that recycles the cash all of this throws off. Miss any one of the three and the story doesn't hold together.

Here's the five-year spine, straight from the filings.

Fiscal year	Period end	Revenue ($M)	Gross margin	SG&A % rev	Net margin	OCF ($M)	Diluted EPS
FY2021	Jan 29, 2022	411.5	53.0%	40.6%	11.5%	28.1	$2.93
FY2022	Jan 28, 2023	467.9	52.5%	39.3%	10.3%	47.3	$3.15
FY2023	Feb 3, 2024	486.1	54.4%	40.9%	10.9%	64.3	$3.65
FY2024	Feb 1, 2025	496.4	54.9%	41.5%	10.4%	47.1	$3.80
FY2025	Jan 31, 2026	529.8	55.8%	43.3%	9.9%	65.1	$3.99

Read that table left to right and the plot appears immediately. Revenue and gross margin both climb every year. Net margin doesn't. That divergence is the whole teardown.

The 55% gross margin hiding in plain sight

Build-A-Bear's gross margin reached 55.8% in FY2025, up from 53.0% in FY2021. That is roughly 280 basis points of expansion in five years, and it puts the company 15 to 20 points above the 35-40% gross margin that's typical for specialty retail, per CSIMarket's industry profitability data. For a business selling something as commoditized-sounding as a stuffed animal, that's a genuinely strange number.

The explanation is that Build-A-Bear doesn't sell a stuffed animal. It sells a birthday party, a school-holiday outing, a grandparent-and-grandkid afternoon. When the purchase is an occasion, price elasticity mostly disappears, because the customer isn't comparing your bear to the one at Walmart. They're comparing the memory to no memory. That's the single most important idea in this entire teardown, and it's why the margin holds.

When I talk to founders about pricing power, this is the fault line I draw. A commodity purchase gets price-shopped every single time, so your margin is set by the market. An occasion purchase gets bought for what it means, so your margin is set by the strength of your brand and the moment you own. The pattern we see again and again is that operators underprice occasion products because they benchmark against commodity competitors who were never really their competition.

The other side of that premium is complexity. One operator running a similar experience-plus-product model put it to us plainly: what's hard about a company like this, when you carry thousands of SKUs, is that it drives up safety stock and it's inherently a bit inefficient, yet customers come precisely because they can get everything they want in one place. You're balancing on a knife edge. Build-A-Bear lives on exactly that edge. Its inventory sat at $82.2M at FY2025 year-end against $529.8M of revenue, and managing that breadth without bloating working capital is the unglamorous discipline behind the headline margin.

The commercial and franchise flywheel

The most underappreciated part of Build-A-Bear's model is the layer that barely shows up in the revenue headline. In FY2025, commercial and international franchising revenue grew 21.6% to $43.9M, which is 8.3% of total revenue, up from $36.1M and 7.3% a year earlier. Core net retail sales grew 5.6% over the same period. So the small segment is compounding about four times faster than the big one.

Segment	FY2024 revenue	FY2024 % of total	FY2025 revenue	FY2025 % of total	YoY growth
Net retail sales (stores + e-commerce)	$460.3M	92.7%	$486.0M	91.7%	+5.6%
Commercial + international franchising	$36.1M	7.3%	$43.9M	8.3%	+21.6%
Total	$496.4M	100%	$529.8M	100%	+6.7%

Why does this matter more than its size suggests? Because it's asset-light. A partner-operated location inside a zoo, a resort, or a theme park is funded by the partner, not by Build-A-Bear. The company supplies the brand, the merchandise, and the format, and collects revenue without carrying the lease or the store labor. International franchise stores work the same way, monetized through royalties, development fees, and merchandise sold to the franchisee. Build-A-Bear has been reported to run partner-operated economics well above what a company-managed store can produce after corporate overhead is allocated, precisely because the fixed-cost burden sits on someone else's balance sheet.

You can see the footprint shift in the store count. Company-managed stores are growing slowly, while partner and franchise locations are exploding.

This is the layer DTC operators most often overlook. When we've watched brands try to grow purely through owned channels, the capex and the customer-acquisition cost climb in lockstep with revenue, and the model never gets easier. A wholesale-heavy operator once made the counterintuitive case to us that contribution margins through retail partners are better than most people assume, that you can genuinely pull 30 to 40% through the right retail relationship, and that the DTC-is-always-better crowd forgets you have to re-acquire the customer every single time you sell direct. Build-A-Bear's franchise flywheel is that same insight, industrialized: let a partner carry the fixed cost, and take the margin. It's a different route to a licensed-IP business than the wholesale-heavy collectibles model we broke down in our Funko teardown, where scale came with real debt on the balance sheet.

SG&A: the shadow over the margin story

Now the uncomfortable part. All that gross margin expansion has a shadow, and the shadow is SG&A. Selling, general and administrative expense rose from 40.6% of revenue in FY2021 to 43.3% in FY2025, about 270 basis points. Compare that to the roughly 280 basis points of gross margin gain over the same window and the math is brutal: the overhead line absorbed almost the entire improvement before it could reach the bottom line.

The proof is in the net margin. Despite five years of record revenue and steadily rising gross margin, net margin drifted from 11.5% in FY2021 down to 9.9% in FY2025. The business is running hard just to stay in roughly the same place on profitability. Store expansion, international buildout, licensing and entertainment investment, and CRM spend are all landing in SG&A, and each is defensible on its own, but collectively they're eating the margin story.

Fiscal year	Gross margin	SG&A % rev	Net margin
FY2021	53.0%	40.6%	11.5%
FY2022	52.5%	39.3%	10.3%
FY2023	54.4%	40.9%	10.9%
FY2024	54.9%	41.5%	10.4%
FY2025	55.8%	43.3%	9.9%

This is the trap I flag most often for growing brands. A rising gross margin feels like winning, and it shows up in every board deck, but it's only real if it survives the trip down to net income. The open question for Build-A-Bear is whether the high-margin franchise and commercial layer can eventually scale enough to pull SG&A down as a share of revenue at the consolidated level. If it can, net margin re-expands. If it can't, the company stays on the treadmill.

Capital allocation: $115M returned, zero debt

The last engine is what Build-A-Bear does with the cash the first two engines produce. Since 2021 the company has returned more than $115M to shareholders through buybacks and dividends, on a revenue base that only just crossed $500M. That included a special dividend, a recurring quarterly dividend, and ongoing repurchases, backed by a $100M buyback authorization approved in September 2024 with most of it still outstanding at last fiscal year-end.

The quieter and more remarkable fact is the balance sheet. Build-A-Bear carried zero long-term debt in every one of the five years in this teardown. No interest expense, no covenant risk, no refinancing exposure on a retail business that inherently carries fixed lease obligations. Operating cash flow reached $65.1M in FY2025. When a business generates more free cash than it needs to grow and refuses to lever up, capital allocation becomes a genuine strategic weapon rather than a survival exercise.

The Build-A-Bear teardown is really one sentence: a strong enough brand can charge an occasion premium that produces a 55% gross margin and enough free cash to fund its own growth, but that gross margin only matters if you keep SG&A from eating it on the way to net income.

What the Build-A-Bear model teaches DTC operators

Strip out the plush and here's what transfers to a $5M-to-$150M DTC brand.

First, experience is pricing insulation. If your product is genuinely an occasion purchase, you can defend a premium margin that a commodity competitor cannot touch, because your customer isn't running a price comparison. The strategic move is to figure out honestly which side of that line you're on, and if you're on the occasion side, to stop underpricing against competitors who were never your real competition.

Second, an asset-light expansion layer changes the growth math. Build-A-Bear grows locations without proportional capex by letting partners and franchisees carry the fixed cost. Your version might be wholesale, marketplaces, or licensing, but the principle holds: not every unit of growth has to sit on your balance sheet and consume your cash.

Third, and most important, watch the net line, not the headline margin. Build-A-Bear expanded gross margin nearly 280 basis points and still watched net margin slip, because SG&A grew just as fast. When we sit down with operators celebrating a gross-margin win, the first question is always whether it survived to the bottom of the P&L. Usually it didn't, and nobody had noticed. That gap between headline margin and net margin is exactly the kind of thing an interim CFO is built to catch before it compounds for five years.

The one question this whole teardown forces: is your product a commodity purchase or an occasion purchase? Answer that honestly and most of your pricing, channel, and margin strategy falls out of it.

Related reading. For another look at how a consumer brand runs the same P&L math, see the Carter's teardown and the Lovesac teardown. For how we help brands model margin and cash, see our fractional CFO work.

Sources and methodology

Primary financial data from SEC EDGAR. All revenue, gross margin, SG&A, net margin, operating cash flow, diluted EPS, inventory, and long-term-debt figures for FY2021 through FY2025 come from Build-A-Bear Workshop's machine-readable XBRL company facts filed with the SEC (CIK 1113809). Build-A-Bear's fiscal year ends in late January or early February; FY2021 ended January 29, 2022, and FY2025 ended January 31, 2026. See the SEC EDGAR company facts API and the Build-A-Bear EDGAR filing index.

Segment and store-count detail from company disclosures. The net retail versus commercial-and-franchising split and the 21.6% segment growth are drawn from Build-A-Bear's FY2025 Form 10-K income statement (net retail sales $486.0M, commercial $38.8M, international franchising $5.1M) and its FY2025 results release, reported March 12, 2026. This teardown combines commercial and international franchising into one asset-light figure. The interim-quarter store-count breakdown by type comes from third-party trackers of the company's disclosures and is point-in-time; the FY2025 10-K reports 662 total year-end locations (375 corporately-managed, 178 partner-operated, and the balance franchise).

Industry benchmark. The 35-40% specialty-retail gross-margin comparison comes from CSIMarket's specialty-retail industry profitability data, used only as a category reference point, not a company-specific comparison.

Operator commentary. The founder-voice observations are drawn from anonymized operating conversations with DTC and retail brand leaders and are presented without any identifying detail. They illustrate patterns, not any single company.

A note on operating margin. The XBRL spine does not report a clean operating-income line for these years, so this teardown works from gross margin, SG&A as a percentage of revenue, and net margin rather than a stated operating margin. Any operating-margin figure elsewhere is typically derived from gross profit minus SG&A and excludes other income and expense.

Frequently asked questions

how does build-a-bear make money if it's a mall brand and foot traffic is falling?

It sells an occasion, not a product. A birthday-party build isn't price-shopped against a stuffed animal at Walmart, so Build-A-Bear can hold a premium. In FY2025 it ran a 55.8% gross margin and $529.8M in revenue, most of it from company-operated stores, plus a fast-growing franchise and commercial layer.

what is build-a-bear's gross margin and how does it compare to other retailers?

55.8% in FY2025, up from 53.0% in FY2021. Typical specialty retail sits at 35-40%, so Build-A-Bear runs 15-20 points above the category. The gap comes from occasion pricing and a heavy mix of experience revenue that customers don't comparison-shop.

why is build-a-bear's franchise and commercial segment growing so fast?

Because it's asset-light. Partner-operated and franchise stores are funded by the partner, not Build-A-Bear, so the company earns royalties, development fees, and merchandise sales without carrying the lease or labor. That segment grew 21.6% in FY2025 versus 5.6% for core retail.

how did build-a-bear return over $115M to shareholders with zero debt?

Free cash flow. A high-margin, occasion-driven business throws off more cash than it needs to grow, and Build-A-Bear ran zero long-term debt every year from FY2021 to FY2025. That cash went to buybacks, a recurring dividend, and a special dividend rather than to interest payments.

why is build-a-bear's sga creeping up even as gross margin improves?

Store expansion, international buildout, and licensing and entertainment investment all hit the SG&A line. SG&A rose from 40.6% to 43.3% of revenue over five years, which absorbed almost the entire gross margin gain. It's the single biggest risk in the model right now.

what can a dtc brand actually learn from build-a-bear's numbers?

Three things. Experience insulates price, so if your product is an occasion purchase you can defend a premium margin. An asset-light expansion layer lets you grow without proportional capex. And gross margin gains mean nothing if SG&A grows just as fast, so watch the net line, not the headline margin.

is build-a-bear's declining e-commerce a problem?

Not necessarily. Consolidated e-commerce demand fell 5.5% in FY2025, but total revenue still hit a record. The in-store experience drives higher average order value and attach rates than online commodity gifting, so the mix shift toward stores is likely margin-positive rather than a warning sign.