The old engine still works. MONDAY.COM IS REBUILDING IT ANYWAY.

The quarter is the easy part to admire. In the three months to the end of June, monday.com brought in $364.6 million, up 22 percent on the year. It earned $1.48 a share against the $1.11 or so that analysts had penciled in, and its revenue cleared their mark as well. Operating income landed at $61.1 million, a 17 percent margin, a few points fatter than the company had promised for the year. The number of customers paying it more than half a million dollars a year grew by 68 percent. And the revenue already signed but not yet counted climbed 34 percent, to $937 million.

That is not the ledger of a company in trouble. It is the ledger of a company running with the wind at its back.

Its leaders knew it, too. The co-chief executives, Roy Mann and Eran Zinman, told investors the quarter had reinforced their conviction that the strategy was working and that it was time to move faster. Their finance chief, Eliran Glazer, said the results demonstrate the underlying strength of the business. Hold that thought for a second. A company that has just proven how strong it is should be the last one reaching for the knife. This one was already cutting.

Rewind three weeks, to before any of those numbers were public. monday.com filed a notice with the Securities and Exchange Commission and told the market it would let go of about 20 percent of its people, somewhere near 600 to 630 jobs by outside counts. You know the usual shape of this story. A company stumbles, misses its targets, and cuts to stay alive. That is not this story, and monday said so plainly. It called the layoff not a rescue but a renovation: a way to align its organizational structure with its strategic focus on the AI Work Platform, part of a wider shift of its product, marketing, and sales toward what it called a leaner, more focused operating model. It even said it would keep hiring where the work mattered.

So read the order of events, not the headline. A profitable, growing company chose to part with a fifth of its workforce, and it framed the reason as design, not distress.

The finances say the same thing. The cuts will cost $45 to $55 million to carry out, and monday is paying that bill while raising its profit forecast, not trimming it, lifting the full-year margin outlook from around 13 percent toward 15, and by the time the quarter was reported, toward 16. This was not a company plugging a hole. It was a company buying a renovation, and turning a profit as it swung the hammer.

There is no more expensive thing a management team can say than a 20 percent layoff. It is the hardest move to reverse and the one the market reads most closely. monday spent that kind of signal not to paper over a miss, but to push its people and its money toward a version of itself it does not yet run at scale.

To see what monday is building, you have to remember how it has always made its money, because the new idea only makes sense against the old one.

For its whole life, the company has grown an account one way: by adding people to it. More seats, more upgrades, more teams signing on inside the same customer. The software was priced by the head, so growing inside an account was, underneath it all, a bet that the customer would keep hiring. It worked. It is most of the reason a startup turned into a company this size.

That is the engine monday has just opened up. For anyone who joins its AI Work Platform after May 6, what they buy is no longer only seats. It is seats plus AI credits. The company puts it simply: seats cover the people using the platform, while AI credits cover supported AI usage. Credits are fuel. When the software's AI agents go to work, they burn credits, and when a customer wants more work done, they buy more. The help page does not hide the point: a customer can add capacity without changing their seat count.

Stay on that last phrase, because the whole redesign is folded inside it. Under the old model, an account grew mainly when the customer grew and hired. Under the new one, an account can grow simply because the customer is putting its machines to work, even if not one more human ever logs in. Growth used to mean people. Now it can also mean output.

Zinman described the shift like a man watching it happen on a screen in front of him. Before the changes in the product and the pricing, he told investors, the only way customers could expand was to add more people and more seats. Now, for the first time since the company shipped its AI agent, he said, customers are expanding not only on the seats for humans, but on AI consumption. Mann, his co-chief executive, said it the way an operator would. Customers top up and reach the end of their consumption bucket and then add more, which, for him, is the best indication that they get value and want more of it.

The figures behind this are still small, and worth naming as small. Revenue from AI products doubled from the first quarter to the second, and it made up 17 percent of the new recurring revenue monday booked in the quarter. That is not a second engine yet. It is the first cough of one turning over.

Changing the price is a decision about economics. What monday did next was a decision about people, and it pointed the same direction.

The layoff did not fall evenly. Casey George, the company's chief revenue officer, told investors the cuts to the sales organization were primarily focused around non-quota carriers and downmarket resources, and that clearing them would afford the company the opportunity to accelerate its investment upmarket. In plainer words: George thinned the ranks that chase smaller customers and the roles that carry no sales quota, and he moved that weight toward bigger accounts, and toward a newer job, putting engineers next to customers to get the AI actually switched on.

That is a company changing not only how it charges, but who it sells to and how. The engineer sent to sit beside the customer is a tell all on its own. Plenty of customers want to adopt AI, Zinman said, but a lot of them do not know how to do it, and helping them deploy products has become, in his words, very strategic for the company. Selling someone a license is one job. Standing up a working AI system inside a large enterprise is another job entirely, and monday is now staffing for the second.

It went one level deeper, at the product line itself. The company is sharpening its focus on monday service and monday CRM, Zinman said, and each will operate with its own dedicated product development, its own go-to-market motion, and its own investment road map. A company that still believes in one motion for everything does not break its products into separate ones. It does that only once it has decided that different buyers now need different machines.

None of this arrived as four separate memos, but it should be read as one, because that is how it happened. Price, headcount, sales coverage, product structure, all of it moved in a single quarter, and all of it moved the same way: away from the broad, seat-led, downmarket motion that built the company, and toward a narrower, consumption-aware, enterprise-led one.

Now for the number almost everyone scrolled past on the way to the earnings beat, which happens to be the one that matters most.

It is called net dollar retention, and it asks a simple question. Take the customers you already had a year ago, and look at whether they are paying you more or less today. Anything over 100 percent means the base you already own is growing on its own, no new logos required. For a software company, it is about the clearest read there is on whether the growth machine still turns without being fed.

monday's net dollar retention came in at 109 percent. Line up a few years and the slope is gentle but unmistakable: 112 percent across 2024, 110 across 2025, 109 now, with the company guiding to 108 by year end. Glazer put the softness down to lapping prior pricing actions taken in fiscal years 2024 and 2025, and to upgrades and cross-selling coming in slightly below the company's original expectations.

The easy reading almost writes itself. Retention is falling, and falling is bad. But that is not how we read it, and it is not what the evidence says. Customers are staying. Gross retention is strong. Enterprise seats keep growing. This is not the chart of a business coming apart at the seams.

It is the chart of a business whose engine is being rebuilt while the old dial keeps watching. Net dollar retention was invented for a seat-led world, where an account grew because it added people. monday is now, on purpose, building a second way to grow an account that the seat-era version of that number was never made to catch. The number is not broken. The business has simply outgrown what the number can see.

That is why the drop is so easy to shrug off, and why it matters anyway. A metric losing a point a year does not look like news. It looks like weather. But a gauge built to measure one kind of growth will read low exactly while a company shifts to another kind, and it will keep reading low until the new engine is big enough to move the needle. The gauge is not failing. It is just early.

And management came close to saying so outright, in the one choice from the quarter worth studying hardest: it declined to raise its full-year revenue guidance to bank the beat. Glazer did not dodge the reason. The company was watching the near-term cost of the 20 percent workforce reduction, he said, and it did not want to lay an aggressive top line raise on top of execution risk that had not fully played out yet. The market caught the caution and not the beat, and the stock slipped on the guidance rather than the quarter.

Do not read that as the market losing faith in the business. Read it as the people with the most information refusing to treat one good quarter as proof that the hard part was over. When the insiders will not bank the beat, the beat is not the most important thing they just told you.

Set the two halves of the quarter side by side one more time.

Revenue up 22 percent. Margins wider. The biggest customers growing fastest. Customers staying put. By every measure that looks backward, the machine monday has run for years is still humming. And right in the middle of that hum, the company let go of a fifth of its staff, rebuilt its pricing around machine usage, marched its sales force upmarket, split its products into separate teams, and warned that its headline growth number would keep drifting down. The old engine still works. monday is rebuilding it anyway.

The move deserves a name, so here is one: elective re-architecture, the choice to redesign a working revenue system before the old one has visibly failed. It is the harder call, and the more telling one. A company forced to rebuild something broken is only reacting. A company that rebuilds something that still works has decided where the world is going and chosen to move first, while moving is still cheap. It is not a new trick. Adobe tore up a profitable licensing business and rebuilt it around subscriptions while the licenses were still selling, and that looked reckless right up until it looked obvious.

To be clear, the judgment here is not that the seat model is dead. It clearly is not. Enterprise seats are still growing, and monday said as much. The judgment is narrower and firmer: the company has decided that seats alone will not carry the phase it is walking into, and it is spending real, hard-to-reverse money to prove itself right before the numbers ever force the question.

That is what makes this worth watching rather than filing away. AI is loosening a grip the human seat has held for twenty years on how software companies grow inside an account. Once a machine can do more work without more people signing in, growth stops being a headcount story. Every company whose growth math was built on seats is walking toward the same question monday is answering out loud: redesign the pricing, the coverage, and the forecasting for that world now, or wait for the old numbers to make the problem impossible to ignore. monday will not be the last name at this fork. It is just an unusually clean, unusually early example of a company taking the turn while taking it is still a choice.

The fair counterpoint is that monday could get the order wrong. The new engine might not scale before the old one slows. Usage-based revenue might prove lumpier than seats ever were. The push upmarket might starve the smaller base that still pays a lot of the bills. Those are real risks, and the softening retention line and the careful guidance are exactly the dials where they would show up first. But look at what kind of risks they are. Every one is a risk of execution, not of direction. The direction is settled. The bet is already down, in cash, in people, and in things that cannot easily be undone.

monday.com might run this transition beautifully. That is not the signal, and it is not our claim. Our claim is quieter and harder to shake: the transition is already underway, paid for out of a quarter the company won, and the instruments most people are watching will be the last to show it.

Revenue grew. Margins improved. Earnings beat. Not one of those will tell you whether the rebuild worked, because every one of them measures the engine monday has already decided is not enough. The question is no longer whether the company is changing its revenue architecture. It is whether the new one compounds before the old one runs out of road. That race is the whole story now, and the strong quarter that opened this piece is the least interesting thing in it.

So, one last time, because it is the point. We do not report the signal. We read it. And here is our reading: the scoreboard said monday won, and monday, looking at the very same scoreboard, decided that winning and being built for what comes next were not the same thing.