Sweetgreen posts a $125.8M "profit" while same-store sales crater 12.8%

---

There is a particular kind of earnings report that demands you read it twice, because the first reading tells you the opposite of the truth. Sweetgreen's first quarter of fiscal 2026, reported May 7 for the period ended March 29, is that kind of report. The top of the press release carries a number the company has not been able to print since it went public in November 2021: net income, and not a sliver of it — $125.8 million. For a chain that has lost money every full year of its existence as a public company, that figure reads like vindication. The growth story worked. The patience paid off.

It did not. The $125.8 million is, almost in its entirety, the proceeds of an accounting event. In the quarter, Sweetgreen recorded a gain of $160.6 million on the disposal of its Spyce business — the robotics startup it acquired in 2021 and the technological seed of its automation ambitions. Sell an asset for a large book gain and it flows straight through net income, regardless of whether you sold one more salad than the year before. Take that gain out, and the picture that remains is not a company turning a corner. It is a company sliding down a slope. Loss from operations was $34.3 million, wider than the $28.5 million loss a year earlier. Adjusted EBITDA — management's own preferred measure of cash earnings power — was negative $8.1 million, a swing of nearly $8.4 million from the roughly break-even $285,000 it posted in the prior-year quarter. The "profit" and the business are moving in opposite directions, and only one of them is real.

This is the forensic problem with Sweetgreen at this moment: the most prominent number in the quarter is the least informative. To understand the company you have to do the thing the headline discourages — look underneath it. And underneath it, the story is one of demand evaporating faster than almost any operating fix can compensate, of a capital-heavy automation bet being deployed into a shrinking sales base, and of a market that has already voted, marking the stock down roughly 90% from its IPO price. The bull case is not dead, and we will give it a fair and specific hearing. But it now has to be argued over a quarter in which traffic fell 11.2% in twelve months. That is a very deep hole to climb out of.

The profit that isn't: quality-of-earnings and the Spyce mirage

Start with the number everyone will quote and almost no one will dissect. Net income of $125.8 million includes a $160.6 million gain on the disposal of the Spyce business. Arithmetic alone tells you that without that gain, the bottom line is deeply negative — the operating loss was $34.3 million, and a one-time gain does not change the cost of running a restaurant. This is the textbook definition of low-quality earnings: a profit driven not by selling more of your product at better margins, but by a non-recurring transaction that will never repeat. Next year there is no second Spyce to sell.

What makes the maneuver more pointed is what Spyce was. Spyce was the robotics company Sweetgreen bought to build the Infinite Kitchen — the automated assembly line that is the centerpiece of management's efficiency narrative. The company is simultaneously telling investors that automation is its future and booking a $160.6 million gain for parting with the asset that originated that automation. There may be a perfectly sound operational reason — perhaps the technology has been internalized and the standalone entity was no longer needed — but the optics are awkward and the accounting consequence is unmistakable: a struggling operating quarter wrapped in a headline of GAAP profitability that the underlying business did not earn.

The honest scorecard for the quarter, then, reads: revenue down 2.9% to $161.5 million; loss from operations of $34.3 million, with the operating margin deteriorating to negative 21.3% from negative 17.2% a year earlier; adjusted EBITDA of negative $8.1 million; and a restaurant-level profit margin of just 10.0%. Every one of those metrics is worse than the comparable period a year ago. The single metric that improved — net income — improved for a reason that has nothing to do with the restaurants. When a company's only positive number is the one it can't repeat, the burden of proof shifts hard onto the bulls.

A 12.8% same-store collapse is not a "deceleration"

Language matters in earnings analysis, and the kindest available word for Sweetgreen's traffic problem is being overworked. Same-store sales did not "decelerate" in the first quarter. They fell off a cliff. Comparable sales declined 12.8%, against a 3.1% decline in the prior-year period — meaning the rate of contraction quadrupled year over year. This is not a growth story losing a step; it is a sales base in active retreat.

The composition is the alarming part. The 12.8% decline broke down into an 11.2% drop in transactions — actual customers walking through the door — a 2.3% drag from product mix, partially offset by a meager 0.7% benefit from menu price increases. Read that again. The overwhelming majority of the damage was traffic. Sweetgreen is not suffering from a pricing mistake it can reverse with a promotion; it is suffering from people choosing not to come. More than one in nine of last year's customers, in aggregate, simply weren't there this quarter.

Management attributes the softness to two factors: adverse winter weather in the current period, and a tough comparison against the prior-year launch of Ripple Fries, which had juiced first-quarter 2025 traffic. Both are real and both deserve weight — weather is genuinely exogenous, and lapping a successful product launch is a legitimate headwind. But neither is a complete explanation, and the magnitude should give pause. A weather quarter and a tough comp might cost a healthy chain a few points of comp. An 11.2% traffic decline is the kind of number that suggests the issue is not only the calendar but the value proposition: a salad that, for many consumers, has drifted to the upper bound of what a fast-casual lunch can credibly cost.

Cyclical demand priced — and excused — as if it were secular weather

There is a frame restaurant analysts use to separate temporary pain from structural decline: is the demand drop cyclical, something that snaps back, or secular, a permanent shift in who eats here and how often? Sweetgreen's narrative leans entirely on the cyclical reading — bad weather, hard comp, transient. The forensic question is whether the evidence actually supports that comfort.

Consider the customer. Sweetgreen sells a premium, discretionary lunch concentrated in dense urban markets and office corridors. That is precisely the cohort most exposed to two slow-grinding forces: the long tail of hybrid work, which thinned the weekday office-lunch crowd that Sweetgreen's economics were built around, and a consumer who, after years of cumulative food inflation, has grown sharply more selective about a $15-plus salad. When demand for a discretionary product falls 11.2% on traffic in a single year, the burden is on management to prove the cause is the thermometer rather than the structure — and a single quarter of weather commentary does not discharge that burden.

The tell is in the guidance, which we examine next. A company confident its Q1 collapse was a weather blip would guide to a sharp recovery. Sweetgreen did not. It guided the full year to a same-store sales decline of 4% to 2% — better than the brutal first quarter, yes, but still negative for the entire year. Management's own forecast, in other words, concedes that 2026 is a year of shrinking comparable sales. That is not what a cyclical-snapback looks like on a spreadsheet. It is what a business resetting to a lower demand baseline looks like.

The guidance quietly admits the recovery is slow and shallow

Guidance is where management's private view leaks into public numbers, and Sweetgreen's full-year 2026 framework is more revealing than its tone. Three figures define it. Same-store sales: a decline of 4% to 2%. Restaurant-level profit margin: 14.2% to 14.7%. Adjusted EBITDA: $1 million to $6 million.

Take them in order. A full-year comp decline of 2% to 4%, after a 12.8% first-quarter drop, mathematically requires the remaining three quarters to improve substantially — but it still lands the year in negative territory. The company is guiding to a second consecutive year of comparable-sales contraction. For a "growth" stock, that is a quiet but profound concession: the engine of the equity story, same-store sales momentum, is not expected to run forward this year. It is expected to run backward, just less violently than it began.

Now the adjusted EBITDA range: $1 million to $6 million for the entire fiscal year. On revenue that will run well over $600 million, an EBITDA midpoint near $3.5 million is a margin so thin it rounds to zero. This is a company whose best case for the year is to generate a few million dollars of adjusted cash earnings — before the heavy capital spending that the Infinite Kitchen rollout demands. And remember, adjusted EBITDA flatters the picture: it adds back stock-based compensation and other real costs. The GAAP operating loss for the year will be far larger. A business guiding to barely-positive adjusted EBITDA, while still posting GAAP operating losses, four-plus years after its IPO, is not a maturing growth story. It is a story still searching for the unit economics that justify its existence.

Infinite Kitchen: demonstration economics, deployment risk

The bull case's trump card is the Infinite Kitchen — the automated assembly system that, in the locations where it runs, is said to cut labor, improve order accuracy, and lift restaurant-level margins. Management has been clear that automation is the path to the profitability the chain has never reached. It is a genuinely interesting bet. It is also, at this moment, a bet being placed into a falling market, and that combination carries a specific risk that deserves naming.

Automation is a fixed-cost solution. You spend the capital up front to retrofit or build a restaurant with the robotic line, and you earn it back through labor savings per transaction over time. The entire return math depends on throughput — on volume flowing across that expensive line. Now overlay an 11.2% traffic decline. When transactions fall, a fixed-cost automation investment de-levers: you've spent the capital, but there are fewer orders over which to amortize it. The Infinite Kitchen's payback period stretches precisely when sales shrink. The technology can be working exactly as designed at the unit level and still be a poor capital allocation if it is being deployed into restaurants whose customer counts are dropping.

This is the gap between demonstration and deployment. A handful of Infinite Kitchens showcasing strong margins in favorable locations proves the concept can work. It does not prove that systematically deploying capital-intensive automation across a fleet — for fiscal 2026 the plan is roughly 13 net new restaurants, about half with Infinite Kitchen — generates an attractive return when the comparable-sales backdrop is negative. The company is asking investors to fund a capital-heavy rollout on the promise of throughput it is currently losing. Until the comp line turns positive, the automation thesis is being underwritten on faith that the denominator — transactions — will recover. The Q1 data points the other way.

The denominator problem in restaurant-level margin

There is a subtler trap in the restaurant-level profit margin, the metric bulls cite most to argue the core stores are healthy. In the first quarter that margin was 10.0%, on $16.2 million of restaurant-level profit. The figure sounds reassuring until you remember what drives it and what it omits.

Restaurant-level profit margin is computed before corporate overhead, before stock-based compensation, before depreciation on the very capital — including the Infinite Kitchens — that management touts. It is, deliberately, a flattering line. And it is under direct pressure from the traffic collapse: when transactions fall 11.2%, the fixed components of store-level cost (rent, base labor, equipment depreciation) spread across fewer orders, compressing the margin. That is why guidance for the full-year restaurant-level margin sits at 14.2% to 14.7% — management expects the metric to improve from Q1's depressed 10.0% as seasonality and easier comps arrive, but the first quarter shows how quickly the number erodes when the denominator shrinks.

The cleaner way to see it: even at a "healthy" double-digit restaurant-level margin, the consolidated company lost $34.3 million from operations and burned cash on an adjusted-EBITDA basis. The store-level metric can look fine while the enterprise loses money, because the store-level metric is engineered to exclude the costs that make the enterprise unprofitable. Investors who anchor on restaurant-level margin are looking at the one number least connected to whether Sweetgreen, the public company, makes money. It does not.

The market has already rendered a verdict

Markets are imperfect, but a roughly 90% drawdown is a loud signal worth taking seriously rather than dismissing. Sweetgreen priced its IPO at $53.00 in November 2021. By early 2026 the shares changed hands around $6, with a 52-week range running as low as $5.00. A $1,000 investment at the IPO would have lost the overwhelming majority of its value over four years. The market capitalization, recently around $700 million, is a fraction of the company's post-IPO peak.

A skeptic of short theses will say the decline is the opportunity — that all the bad news is priced, that a chain trading near all-time lows offers asymmetry to the upside. That argument has merit and we address it squarely below. But the forensic point here is different: the price collapse is not noise to be argued away. It is the cumulative judgment of years of investors watching a company promise profitability and deliver losses, quarter after quarter. The analyst community reflects the same caution — the average twelve-month price target sits in the high-single digits, with at least one firm carrying a Neutral rating around $7.50. Wall Street is not modeling a turnaround; it is modeling a company that might, eventually, stabilize. When the consensus target barely exceeds the current price after a 90% fall, the asymmetry the bulls invoke is far less obvious than it sounds.

Consumer-discretionary exposure at the worst possible point in the cycle

Sweetgreen's vulnerability is not only company-specific; it is structural to what it sells and to whom. A premium salad is among the most discretionary lunch purchases in fast casual. It competes not only with CAVA and Chipotle but with the cheaper, more defensive option every consumer always has: bringing lunch from home, or trading down to a fast-food value menu. In a consumer environment where households have spent years absorbing cumulative food inflation, the $15-plus bowl is exactly the line item that gets cut first when budgets tighten.

That is why the 11.2% traffic decline should be read in macro context as well as company context. It is consistent with a discretionary consumer pulling back precisely at the premium end where Sweetgreen lives. And it compounds the automation problem: a company making a fixed-cost capital bet needs volume certainty to justify the spend, but its product sits in the part of the consumer wallet with the least volume certainty. The cyclical exposure and the capital intensity are not two separate risks. They multiply each other.

What the bulls genuinely get right

A fair forensic case has to concede where the other side is genuinely strong, and on Sweetgreen the bull case has real substance that the bear narrative should not wave away.

First, the balance sheet buys time. Sweetgreen ended the quarter with roughly $156.8 million in cash and equivalents — and that was before the Spyce proceeds, which materially strengthen liquidity. For a company guiding to barely-positive adjusted EBITDA, the absence of an imminent funding cliff matters enormously. This is not a business about to be forced into a dilutive raise or a distressed sale. It has runway to let the automation thesis play out, and runway is the scarcest asset a turnaround can own.

Second, the Infinite Kitchen is a real, differentiated asset, not vaporware. Sweetgreen is genuinely further along on in-store automation than most fast-casual peers, and if the technology delivers the labor savings and throughput gains management describes, it could structurally lift restaurant-level margins in a way competitors cannot easily replicate. Owning the robotics capability — even after divesting the Spyce shell — gives Sweetgreen a credible, ownable efficiency lever. Many money-losing growth stories have no such lever at all.

Third, the digital mix is a genuine strength. Total digital revenue reached 67.2% of sales, up from 59.9%, and owned digital — the highest-value, most data-rich channel — rose to 38.9% from 31.9%. A high and rising owned-digital share is exactly what you want: it lowers third-party marketplace fees, deepens customer data, and supports loyalty economics. That is real progress on the parts of the model management can control.

Fourth, the weather and comp explanations are not fabrications. Winter weather genuinely suppresses restaurant traffic, and lapping the Ripple Fries launch is a real, quantifiable headwind. A meaningful portion of the 12.8% comp decline is almost certainly transient, and the back half of 2026 should show easier comparisons. If traffic stabilizes even modestly while the digital mix and automation margins keep climbing, the path to positive adjusted EBITDA is visible. The bull case is not a fantasy. It is a bet that the structural assets outlast the cyclical pain — and that is a defensible bet for a patient investor who can stomach the volatility.

Where the bull case has to be right

Concede all of that, and the thesis still reduces to a single load-bearing assumption: that the traffic decline is temporary and reverses. Everything else the bulls cite — the cash cushion, the automation lever, the digital mix — only pays off if customers come back. The cash extends runway but does not create demand. The Infinite Kitchen improves unit margins but de-levers if throughput keeps falling. The digital mix optimizes the channel but cannot manufacture transactions that aren't happening. Every strength is conditional on the one thing Q1 showed deteriorating: traffic.

And management's own guidance refuses to underwrite that reversal. A full-year comp guide of negative 2% to negative 4% is not a forecast of recovery; it is a forecast of slower decline. The company is telling investors, in its own numbers, that it does not expect same-store sales to grow this year. For a stock whose entire valuation rests on eventual growth re-accelerating, that is the quiet admission that matters most. The bull case requires the traffic to turn. The guidance says it won't — not in 2026.

The kicker

Strip away the framing and Sweetgreen's first quarter says one thing plainly: the only profit it could show was the profit of selling the future it keeps promising. The $125.8 million net income is the wake of the Spyce sale, not the wake of a recovering restaurant. The real engine — comparable sales — ran backward 12.8%, on an 11.2% traffic decline, into a full-year guide that concedes the line stays negative all year. The cash is real, the automation is real, the digital mix is real; what is not yet real, four and a half years and roughly 90% of its market value after the IPO, is a single quarter of the unprofitable growth company actually growing, or actually profiting, from the thing it sells. Until the denominator turns — until customers come back through the door — every margin metric, every automation slide, and every adjusted figure is an elaborate accounting of throughput the company is currently losing. The bulls may yet be right that the assets outlast the pain. But you cannot automate your way to profit on traffic that keeps walking out, and you cannot sell Spyce twice.

A company that books its largest-ever profit by selling the very robot it calls its future, in the same quarter its customers fall 11.2%, has not found the bottom — it has merely found a buyer for the one asset it could spare.