monday.com Cut 630 Jobs. The Ones With a Quota Stayed.

monday.com has a phrase for the part of the company it cut, and it took an analyst asking twice to get it said out loud.

The company reduced its global workforce by about 20% on July 22, roughly 630 people. For nearly four weeks the public explanation was a leaner operating model and an AI Work Platform. Then, on the August 17 earnings call, Jefferies analyst Brent Thill followed a colleague’s question with a complaint:

“I guess, are you making changes on the go-to-market where you’re reducing the go-to-market team in a material way? Because, again, I think it wasn’t very clear to his question what is actually going on in the go-to-market.”

Chief revenue officer Casey George answered it directly.

“The restructure of the go-to-market organization was primarily focused around non-quota carriers and downmarket resources.”

In the same answer, he said the sales roles that carry a number are growing: “So we expect our headcount for that cohort of our sales team to grow for the year.”

A quota carrier is someone with a revenue number attached to their name. In a go-to-market org, the people without one are marketing, marketing ops, enablement, partner management, support and much of customer success.

That’s the group the cut concentrated on. It was stated by the executive who ran it, on a call where the words carry legal weight.

We wrote four days ago that the companies selling you AI are shrinking their own teams first. This is the next question answered: not which companies, but which desks.

The savings go to the AI bill

The second thing the call surfaced is where the money went, and it isn’t to the bottom line in the way the framing suggests.

Co-CEO Roy Mann described the cut as the hardest decision since founding the company and said “Most of the savings will be reinvested in the people, products and AI.” Chief financial officer Eliran Glazer put a number on it and then took part of it back:

“So we said that the annualized cost savings from a growth perspective is expected to be $100 million. We said we’re going to invest the vast majority of it is going to be into talent, product, AI. Obviously, there is a lot of cost related to AI. So this is something that will reduce, obviously, the savings.”

Read that last sentence again. The company removed roughly 630 salaries, and its finance chief says the running cost of AI will eat into what those salaries freed up.

That’s the trade, stated plainly by the person who has to reconcile it. Headcount on one side, compute on the other. Most companies describe this as efficiency. monday.com described it as a transfer. An analyst asked about margin trajectory, and the honest answer required naming the new costs.

What the numbers actually say

Three figures are worth carrying out of the call, because two of them are commonly misread.

Headcount at the end of Q2 was 3,169, down only 42 from Q1. The 630 reductions are almost entirely still ahead of the reported number, not behind it. Glazer said the company expects headcount “to be down approximately 20% at the end of fiscal year 2026.”

Margin did improve, but less dramatically than the layoff framing implies. Q2 operating margin was 17%, up from 15% a year earlier. Full-year guidance sits at roughly 16%, with a currency drag of 100 to 200 basis points from the shekel. The company also books $45 million to $55 million in restructuring charges against that.

And the AI revenue that justifies the whole move is early. Co-CEO Eran Zinman said AI annual recurring revenue doubled from Q1 to Q2 and now makes up 17% of net new ARR, then added the qualifier most coverage dropped: “This is a meaningful signal, not because the absolute number is large yet, but because the rate of change tells us customers are actively choosing our AI capabilities.”

Not because the absolute number is large yet. The company cut a fifth of its staff ahead of that number, not after it.

The argument against reading this as an AI story

There’s a real case that this is an ordinary go-to-market repositioning wearing AI language, and it’s worth stating at full strength.

George’s answer named downmarket resources alongside non-quota carriers. He said the company will lean on partners to serve small customers, calling it “a much more efficient sale.” That is a move upmarket, which software companies have run for 30 years without any AI involved. Mann described the operating change as fewer management layers and smaller teams with real decision-making authority, which is org design, not automation. Nobody on the call said a model replaced a marketer.

The reason it still reads as an AI story is Glazer’s sentence about costs. A conventional upmarket shift redeploys budget from one set of humans to another. This one redeploys it from humans to a cost line that the CFO flagged as large enough to reduce the savings. The shape of the reallocation is what makes it new, not the vocabulary around it.

What to take from it

If your role doesn’t carry a quota, skip the part where you worry about your job. Go straight to answering the question George was answering.

Find the number your work moves and the distance between your work and that number. Pipeline sourced, retention on a segment, cost per qualified lead, payback period. Roles that can name the number and show the link are quota-adjacent even without a quota. Roles that can only describe activity are what “non-quota carriers” means on a slide.

The 630 people at monday.com did not lose their jobs because AI wrote better copy. They lost them because a public company decided which parts of its go-to-market motion it could point a number at, and moved the rest of the budget to a compute bill it expects to grow. Dentsu made the same math a three-year target two days later.

The phrase to watch in your own company is not “AI strategy.” It’s “non-quota.”