On September 10, a hamburger chain that has been frying patties since 1934 published a number that set crypto Twitter alight: 19 percent same-store sales growth across its franchised locations. The chain, Steak 'n Shake, tied the moment directly and repeatedly to its decision to accept Bitcoin.
What the same announcement did not publish is equally interesting. No dollar volume of Bitcoin processed. No transaction count. No split between tickets settled in satoshis and tickets settled, as they always have been, in dollars. We were handed an output metric and invited to infer a payments story from it, and the invitation ran in exactly one direction, because only one direction flatters the teller.
I have a reflex about this, built from years of staring at dashboards that always look better than the warehouses behind them. When someone shows you the result and hides the input, the input is where the argument lives.
There is an older shape underneath this one, and it is a shape about roots. Roots do not announce themselves. They grow in the dark, and only years later does anyone notice the tree. The 2025 version of the Bitcoin adoption story is heavy on visible fruit — logos on windows, announcements on X, a chief officer with a title nobody has heard of — and conspicuously light on what is happening underneath the soil. That gap is what I want to open up here, not because Steak 'n Shake did anything wrong, but because the gap has quietly become the load-bearing wall in an entire industry's claim to real-world adoption.
Context: the sequence matters more than the headline
In May 2025, Steak 'n Shake began accepting Bitcoin across its United States locations. The brand dates to 1934 and is controlled through Biglari Holdings, a New York Stock Exchange-listed holding company whose portfolio over the years has included Western Sizzlin, Maxim's, First Guard Insurance, and a collection of restaurant franchises with a long tail of underperforming stores. Steak 'n Shake has spent much of the past decade in turnaround mode. That context matters more than any of the payment technicalities, because for a chain in that position a same-store sales recovery is not a curiosity — it is the difference between staying listed and being absorbed.
The rail the company says it uses is the Lightning Network, Bitcoin's channel-based second layer, which routes most activity off the base chain and reconciles periodically to it. The choice is not decorative. A burger and a milkshake is a ten to fifteen dollar ticket, and at 2025 base-layer fee levels a direct Bitcoin transaction would eat a meaningful slice of that ticket. Lightning, or a custodial processor sitting on top of Lightning, is the only arithmetic that closes.
Alongside the payments rollout, the company placed $10 million of treasury capital into Bitcoin as a strategic reserve — the same corporate balance-sheet maneuver that has spread from software companies into the broader market since 2020.
Then came the September 10 numbers: double-digit same-store sales growth, 19 percent at franchise locations, an assertion of outperformance against McDonald's, Taco Bell, and Domino's over the same window, an earlier data point of 11 percent sequential growth in the second quarter accelerating toward 15 percent in the third, and a claimed savings of roughly 50 percent on payment processing fees.
One detail deserves a flag before we go further. The public commentary attributed to a Michael Boes, described in coverage as a Chief MAHA Officer, is attached to an appearance at an event labeled Bitcoin 2026. If the announcement landed in September 2025, that conference could not yet have happened — Bitcoin's main annual gathering falls in late May. Either the event reference is wrong, the personnel reference is garbled, or the sourcing is a whisper chain that lost fidelity on the way out. When a story's smallest checkable detail wobbles, I start looking at how the big ones are anchored. MAHA is also not a title that exists in restaurant or financial services. That does not make the sales figure false. It makes the reporting incomplete, and incompleteness in the easy parts is a warning sign about the hard parts.
Core: unwinding the 50 percent claim
Start with the actual flow of a Lightning payment at a counter, because the abstraction hides the cost.
A customer opens a wallet, whether custodial or self-custodial. The wallet constructs an invoice, the invoice is paid across one or more channels, and the routing happens in fractions of a second. The merchant's side then does something with the received value. That something is the entire story. If the merchant's chosen processor converts to dollars at settlement, the merchant never holds Bitcoin at all — they hold a dollar receivable, netted through a company whose fee schedule, uptime, and compliance posture now sit between the customer and the till. If the merchant self-hosts, they hold the coin, and they inherit custody responsibility, channel liquidity management, and the obligation to price their own conversion decisions.
Steak 'n Shake has not said which of these it runs. That single omission is the hinge on which the fee claim swings.
A United States card transaction typically costs a merchant between 2.5 and 3.5 percent all-in once interchange, network assessments, acquirer markup, terminal costs, PCI compliance, and chargeback exposure are counted. That range is the ceiling of what any alternative rail can save you. You cannot save more than you were spending.
On the Bitcoin side there are three plausible architectures. A self-hosted stack, the BTCPay Server model, can push marginal transaction cost toward zero while stacking up real operational cost in node maintenance, channel liquidity, custody, and per-receipt tax basis tracking. A third-party processor, of the OpenNode or Strike variety, charges roughly half a percent to one percent and absorbs conversion and settlement. A hybrid keeps small tickets auto-converting while retaining a portion of receipts in Bitcoin.
If the chain is running a processor at one percent against a card cost near three, the reduction on the interchange line is real and substantial — roughly two-thirds. But the interchange line is not the whole cost of accepting money. The card processing fee is a ceiling on savings; the true all-in cost of a Bitcoin receipt is a floor that nobody publishes, and the merchant's genuine savings is the gap between them, not the number in the press release.
That floor has components that never appear in a merchant services pitch. In the United States, every Bitcoin receipt is a taxable event. The merchant recognizes ordinary income at fair value on the date of receipt, and any later disposition of that coin generates capital gain or loss measured against that basis. At a ten dollar average ticket, the accounting labor required to track fair value, basis, and disposition per transaction can plausibly exceed the payment savings per transaction. This is not a trivial footnote. It is the reason most small merchants who "accept crypto" in practice accept it through a processor that handles all of it invisibly, at a fee that quietly restores a portion of what the card network was charging.
Then there is the reserve. Ten million dollars of Bitcoin on the balance sheet of a mid-cap restaurant holding company is a mark-to-market instrument, and since the FASB's ASU 2023-08 rules took effect for fiscal years beginning after December 15, 2024, crypto holdings are measured at fair value with changes flowing through net income. A reserve announced as a statement of conviction becomes a quarterly line item that can swing earnings in either direction, visible to every analyst covering the stock.
This is the underreported risk in every merchant-adopts-Bitcoin story: the fee narrative is small and legible, while the balance-sheet narrative is large and invisible until it isn't. A restaurant accepting Bitcoin to shave a point and a half off processing is running a payments strategy. A restaurant holding ten million dollars of Bitcoin while accepting Bitcoin is running a leveraged macro position with a grill attached. Both can be defensible. They should not be described with the same word, and they should not be reported with the same level of scrutiny.
Now the verifiability problem, which I find more interesting than the arithmetic.
Lightning payments are not individually attributable on chain. Channel opens and closes are visible in aggregate, the routing graph is partially observable, but no outside observer can walk up to the base chain and say these four thousand payments belonged to Steak 'n Shake. Merchants using custodial processors add another layer of obfuscation, because processors batch and net before settling. Part of this is a feature — it is the privacy that makes a Bitcoin payment more palatable than a card swipe that broadcasts your location, your basket, and your brand preference to an advertising broker within milliseconds. Part of it is a measurement impossibility.
So the industry has arrived at a strange place. Bitcoin merchant adoption is the only adoption claim in this entire asset class that is structurally unverifiable from the outside, which means every such claim in a bull market has to be discounted by exactly the size of the incentive to make it. We can audit a stablecoin's reserves, a rollup's sequencer revenue, a lending protocol's outstanding debt. We cannot audit whether your neighborhood burger chain did meaningful Bitcoin volume last quarter. That is not a scandal. It is a limitation of the tool, and it should be printed into every headline that follows.
The same limitation makes the causal claim untestable. Same-store sales at a restaurant chain move for a handful of reasons: menu price increases, promotional calendar, traffic recovery, comparison against a weak base period, delivery mix, and occasionally a genuine shift in customer behavior. 2025 was a year of elevated food-away-from-home pricing in the United States. A high-teens same-store gain in that environment can be substantially price rather than volume. The announcement disclosed no traffic figure, no basket-size decomposition, and no Bitcoin penetration rate. Without those three numbers, the claim that Bitcoin drove 19 percent is not a finding. It is a hypothesis wearing a finding's clothing.
I have been inside a version of this movie. In 2021 I helped run twelve onboarding workshops for women entering the NFT space through the community I founded, Decentralized Hearts. Our growth numbers looked extraordinary on a spreadsheet for about two quarters. Then we looked at retention and discovered that a large share of the new members were the same people cycling through with fresh wallets. Nobody lied. The metric was measuring the wrong thing, and the wrong thing was the thing everyone wanted to hear. Restaurant chains are not immune to that dynamic, and neither are the crypto outlets that cover them. When I later mapped the fee stack for a remittance corridor serving Manila — the corridor that convinced me this technology mattered before the word web3 existed — I found the same pattern in the opposite direction: the headline savings were real, and the operational reality absorbed most of them, until the merchants who benefited most were the ones with back offices.
There is a deeper structural point about what accepting Bitcoin actually means in 2025, and it is where my own view of this space has gotten less romantic than it used to be.
If a customer pays in Bitcoin and the merchant's processor converts to dollars within seconds, what the merchant acquired is not Bitcoin. It is a dollar receivable with a Bitcoiny on-ramp and a new intermediary. The chain still lives inside a fiat accounting system. It still chases monthly revenue targets denominated in dollars. It still has a vendor relationship with a payments company whose obligations, uptime, and fee schedule sit between the customer and the register. The rail changed. The power structure mostly did not. This is the same pattern I keep watching on the layers above the base chain: every added abstraction makes a system easier to use and easier to capture, and the capture happens at the exact point where the money is actually held. The same dynamic animates rollup economics — cheap data availability made those rails cheap, and the moment that subsidy decays, the abstraction gets expensive again, and the merchant or user who never noticed the subsidy is the one who pays for its withdrawal.
And there is a philosophical fork here that gets flattened into marketing copy. On one side sits a payment system where the transaction is nobody's business but the two parties'. On the other sits the state-issued digital currency being engineered in parallel, which moves in precisely the opposite direction: total transaction visibility, policy-conditional spendability, and no meaningful ability to opt out. Those are not two variations of the same idea. They are two different ideas about who gets to watch and who gets to spend. A merchant that accepts the private rail, even imperfectly, even through a processor, has taken a small position in the argument. It is worth noticing, and worth not overstating.
Which is why the details matter so much. A rail that is half a percent cheaper in aggregate, but which requires a compliant processor, a fair-value ledger, and a custodial relationship, is emancipatory only for a merchant with the clerical capacity to support it. For a taco stand, the real frontier is not Bitcoin acceptance — it is infrastructure that lets a taco stand accept it without hiring a bookkeeper. That infrastructure barely exists in the United States. Its absence is the actual story that the September 10 announcement walked right past, and it is the thing I would fund if I were allocating capital this cycle.
A word on the market context, because the timing is not neutral. We are reading this in a period where survival has displaced speculation as the operating principle. In a correction, the projects that bleed are the ones whose revenue depended on inflows rather than usage, and the balance sheets that hurt are the ones holding a volatile asset against fixed obligations — payroll, rent, debt covenants. A restaurant group with a Bitcoin reserve and a soft traffic base is a small, domestic illustration of the exact structure that dismantled lending desks in 2022. The reserve is not the problem. A reserve held without a published policy for what happens at minus forty percent is the problem. Winter does not kill orchards. It prunes them, and what fails the pruning was probably never load-bearing. I have not seen Steak 'n Shake's treasury policy. Until it is public, the ten million dollars is an anecdote, not a strategy.
Contrarian: the most probable causal story is not the one being sold
Let me state the counterintuitive position plainly, because the narrative industry will not.
The most probable driver of Steak 'n Shake's sales improvement is a brand and traffic story that has almost nothing to do with the technical properties of the Lightning Network. A ninety-one-year-old chain mid-turnaround generated national earned media worth millions in equivalent paid placement from a payments announcement. That coverage reached an audience that mostly does not own Bitcoin, walked into a store out of curiosity, and paid with a card. The Bitcoin transactions themselves are, in dollar terms, almost certainly a rounding error against the chain's total volume. The adoption was real. The causal claim grafted onto it was the press.
None of this requires anyone to be lying. Companies do what companies do: they choose the flattering narrative when the unflattering alternative — a turnaround that is working, partly on the back of higher menu prices — is harder to sell to shareholders and much harder to sell to crypto Twitter.
And the sting of it is this. Accepting Bitcoin is, for most chains, not an infrastructure decision at all. It is a marketing line item with a technical noun attached. Integration cost has fallen to the point where a franchise operator can add Lightning acceptance in roughly the time it takes to renegotiate a POS contract. That low barrier is genuinely good for the ecosystem, and it is also exactly why there is no moat in it. Jersey Mike's did a version of this in 2024. Starbucks experimented and pulled back. McDonald's, with something near forty percent of the American burger market, has not needed to. One brand's adoption tells you about that brand's marketing strategy. It tells you almost nothing about whether Bitcoin has become money.
The blind spot underneath the whole conversation is bigger than either side acknowledges. Everyone argues about the fee percentage, the privacy properties, the bullishness or bearishness of the announcement. Almost nobody asks the only question that determines whether any of this compounds: what happens to the merchant's Bitcoin exposure after three consecutive down quarters? If the answer is panic-sell the reserve, then the reserve was never a strategy — it was a trade wearing a mission's clothing. If the answer is hold, publish the policy, disclose the basis, then Steak 'n Shake has quietly done something more consequential than any of its peers: it has dragged treasury risk management out of the finance department and into public view, on a restaurant's balance sheet, under SEC reporting rules, in front of customers who will learn the outcome in real time.
There is one more inversion worth naming. In DeFi, we have spent years treating governance-set interest curves as if they were market rates, when many of them are closer to aesthetic choices than to prices discovered by supply and demand. Traditional payments have the opposite pathology: the rates are real, and they are hidden. A merchant never sees the decomposition of the 2.9 percent that leaves their account. Both systems produce opacity about price. Neither of them is fixed by putting a different logo on the terminal.
Takeaway: what to watch, and what the seeds actually look like
Watch the next disclosure for two numbers that would end the ambiguity: Bitcoin payment volume and the percentage of tickets paid in Bitcoin. If they never appear, assume both are small, and treat every future reference to this case as a marketing claim rather than a data point. Watch for an explicit reserve policy — treasury caps, custody arrangements, whether the coin is held directly or parked with a third party, and what the trigger is for selling. Watch the follow-on behavior across the Biglari portfolio and, more importantly, outside it. Replication is the only real confirmation. A single brand's experiment that lands in a news cycle which cannot distinguish adoption from attention is not evidence of anything except a good publicist.
And if you hold Bitcoin and you are reading this hoping a hamburger chain's same-store sales somehow makes your position safer — it doesn't. Not this quarter. The things that determine whether your assets survive are custody, leverage, and whether the rails you depend on stay funded when the subsidy that made them cheap disappears. Processing some unknown quantity of Bitcoin through an unnamed processor, at a chain with a turnaround narrative and a mark-to-market reserve, does not move any of those three levers.
From the ashes of 2022, we planted seeds for 2030. Honesty requires admitting that most seeds look identical to the dirt they were buried in, right up until the moment they don't. Steak 'n Shake might be one of them. It might also be a very well-covered milkshake.
The next time a company tells you that Bitcoin fixed its sales, ask for the transaction count. If they will not give it to you, the number that mattered was never in the press release.