From Aryan Vatsa | Product & Market Analysis
monday.com Seats Plus Credits: AI Hit 17% of New ARR as Retention Slid to 108%
On this page
In Q2 2026, monday.com said AI products made up 17% of its net new ARR, one quarter after it launched a seats plus credits pricing model. In Q3 it reported $36 million of AI ARR, and net dollar retention fell to 108% with management guiding to 107%. So far, monday.com AI credits look like a relabel of existing growth more than an addition to it. The filings show why, and they show what would change that verdict.
Key takeaways
- AI was 17% of monday.com's Q2 net new ARR, then the metric changed. The Q3 release reported AI ARR in dollars, $36 million, so the two quarters cannot be compared on the same basis.
- monday.com's net dollar retention fell for a fourth straight reported quarter. It went from 111% in Q3 2025 to 110%, 109% and 108% across 2026, and the CFO now expects 107%.
- The AI ARR definition includes credits bundled inside plans. Part of the $36 million is subscription revenue relabelled as AI, which the CFO acknowledged by saying AI ARR is "not incremental yet".
- Customer additions slowed sharply between Q2 and Q3. Net new customers with more than 10 users fell from 767 to 182, a 76% drop, while seat expansion was described only in words.
Reader for this post: founders and revenue leaders at SaaS companies weighing a hybrid seat and credit model, because monday.com is the cleanest public test of whether that model expands accounts. Figures last verified 11 November 2026.
The short answer. monday.com AI credits have not yet shown up as extra retention. AI reached 17% of net new ARR in Q2 and $36 million of AI ARR in Q3, yet net dollar retention fell from 109% to 108% and is guided to 107%. Credits are adding a meter, but the filings do not show them adding net revenue on top of seats.
What monday.com changed in May: seats plus credits
monday.com used to grow an account one way. A customer added people, and each person was a seat. Co-CEO Eran Zinman said on the Q2 call that before the change, "the only way customers could expand was to add more people and more seats".
The Q1 2026 release, dated 11 May 2026, announced an AI Work Platform with native agents and a seats plus credits pricing model. The release described the move as a shift to consumption-based pricing. Seats still pay for people. Credits pay for AI work done by agents, the vibe app builder and other AI features.
The bet is simple to state. If agents do work that people used to do, a seat-only company loses revenue as headcount shrinks. A credit meter lets revenue follow the work instead of the people. That is the same hedge examined in Workday's flex credits as a seat hedge, and the same tension covered in how seat compression reprices SaaS.
Why the top-up is the number that matters
A bundled credit allowance is not expansion. It is part of the plan price. Expansion only happens when a customer runs out and buys more. Co-CEO Roy Mann told analysts in August that customers "do reach the end of their allotted quota and then they increase it and top up". He called it "a new vector for us".
He also said the early figures were "still small numbers". That caveat came from the company itself, in the same quarter that produced the 17% headline. I would weight the caveat more heavily than the headline, because a top-up habit takes several renewal cycles to show up in retention.
The Q2 number: AI was 17% of net new ARR
The Q2 2026 release led with one sentence: ARR from AI products doubled from Q1 and represented 17% of net new ARR in Q2. Revenue was $364.6 million, up 22% year over year, and overall net dollar retention was 109%.
Net new ARR is the annual recurring revenue added in a quarter after churn and contraction. So 17% means that roughly 1 dollar in 6 of the quarter's net growth was tagged as AI. For a product launched six weeks into the quarter, that is a fast start.
The release gave no dollar figure for AI ARR. Without the size of net new ARR, a reader cannot convert 17% into dollars. monday.com does not publish total ARR in its quarterly release, so the percentage was the only handle available in August.
What 17% does and does not tell you
A share of net new ARR rises in two ways. AI revenue can grow, or the rest of net new ARR can shrink. In a quarter where non-AI growth slows, the AI share climbs even if AI dollars are flat. The Q2 call suggests some of both were happening.
CFO Eliran Glazer also flagged a headwind on that call. He said "we may see some modest pressure on NDR as we lap prior pricing actions taken in fiscal years 2024 and 2025". He then said NDR would be "around 108%" for the year. That was the before picture: AI rising, retention guided slightly lower, and price increases from earlier years rolling off.
monday.com Q3 2026: $36 million of AI ARR and a new definition
The Q3 2026 release on 9 November reported revenue of $377.5 million, up 19%. It reported AI ARR of $36 million, "including AI credits, agents and vibe". It did not repeat the share of net new ARR.
That switch is the single most important fact for anyone tracking this story. The Q2 metric and the Q3 metric are different objects. A before-and-after read of AI ARR share is not possible from the filings, because the company stopped reporting the share.
| Metric | Q1 2026 | Q2 2026 | Q3 2026 |
|---|---|---|---|
| Revenue | $351.3M | $364.6M | $377.5M |
| Year-over-year growth | 24% | 22% | 19% |
| Net dollar retention, overall | 110% | 109% | 108% |
| NDR, customers above $50k ARR | 116% | 115% | 114% |
| Customers above $100k ARR | 1,844 | 2,019 | 2,127 |
| Customers above $500k ARR | 99 | 114 | 123 |
| AI disclosure | Platform and pricing launched | 17% of net new ARR | $36M AI ARR |
All figures are from monday.com's quarterly results releases filed on Form 6-K. The AI row changes basis each quarter, which is the point of the table. NDR is the company's trailing four-quarter weighted average.
The definition that matters
The Q3 release defines AI ARR to include stand-alone AI products, AI credit packages sold for a subscription term, and "the portion of subscription plans attributable to included AI credits". The last clause is the one to read twice.
Under that definition, a customer who buys a plan with bundled credits generates AI ARR even if they never top up. Some of the $36 million is therefore an allocation of plan revenue that monday.com would have booked anyway. The release does not say how much.
The CFO said as much on the call. Asked whether AI ARR was incremental, Glazer answered: "it's not incremental yet. It's part of our overall ARR." That is an honest answer, and it is the sentence that should anchor every reading of the $36 million.
For scale, I divided $36 million by Q3 revenue annualised. Q3 revenue of $377.5 million times 4 is about $1.51 billion, so AI ARR equals roughly 2.4% of the revenue run rate. That ratio is mine, not the company's, and it uses revenue as a stand-in for the ARR figure monday.com does not publish.
Credit adoption in the $50,000 cohort
Zinman gave one adoption figure on the Q3 call. Of accounts above $50,000 in ARR, "over 30% of those accounts pay for AI credits and use it". He added that the cohort showed about 4.4 times the AI adoption and usage of the rest of the base.
The Q3 release counts 5,038 customers above $50,000 ARR. Over 30% of 5,038 is at least 1,511 accounts paying for credits. That is a meaningful base for a meter that is six months old, and it is concentrated in the segment with the best retention.
Net dollar retention fell four reported quarters in a row
Net dollar retention, or NDR, compares what a group of customers pays now with what the same group paid a year earlier. Above 100% means expansion outran churn. monday.com computes it as a trailing four-quarter weighted average, per Glazer on the Q3 call.
The path is steady and downward. NDR was 111% in Q3 2025, per the Q3 2025 release. It was 110% in Q1 2026, 109% in Q2 and 108% in Q3. On the Q3 call Glazer said "we believe it's going to go down 1% to 107%".
The guidance moved a point in 90 days
In August the CFO pointed to about 108% for the year. In November the outlook was 107%. That is one point of retention lost from the outlook in a single quarter, during the first full quarter of credit selling.
One point sounds small. On a trailing four-quarter average it is not. A one-point fall in a lagging average means the most recent quarter was weaker than the one it replaced by more than one point. The NDR posts in the Q2 filings retention comparison show how rare steady quarterly declines are among peers that held price.
Management named three causes. Zinman said "SMB customer additions were lower and SMB logo churn was higher", and called that "by design". Chief Revenue Officer Casey George cited macro sensitivity in small businesses, longer upmarket buying cycles, and seasonality. None of the three is the credit model. None of the three is evidence for it either.
Do monday.com AI credits add revenue or replace seat expansion?
This is the question the whole pricing change rests on. The honest answer from the filings is that nobody outside the company can tell yet, and the company says it cannot fully tell either.
What management said about seats
On the Q3 call, UBS analyst Taylor McGinnis asked whether credits were replacing seats. Mann replied: "it's obviously hard to say because they have like one behavior with seats". He went on: "it's generally an addition and it's recurring".
George, asked by Wolfe Research whether seats were falling in the $50,000 cohort, said: "We continue to see seat expansion in this cohort". Mann also cited the largest single deployment, which grew to over 87,000 seats.
Those are claims of direction, not size. The releases contain no paid seat count and no seat growth rate. "Generally an addition" from a co-CEO is the most positive sentence available, and it is hedged twice. I take it at face value as a statement of belief, and I do not treat it as evidence.
What the customer counts say
The customer counts are hard numbers, and they moved sharply between Q2 and Q3. I subtracted each quarter-end count from the next to get net additions in each cohort.
The broad cohort of customers with more than 10 users added 767 net customers in Q2 and 182 in Q3. That fits the SMB explanation. The enterprise cohorts slowed as well, though: net additions above $100,000 ARR fell from 175 to 108.
That second fact matters for the credits question. The enterprise cohorts are where credit adoption is highest. If credits were pulling accounts across ARR thresholds faster, the $100,000 and $500,000 counts would be the first place to see it. They slowed instead.
| Signal | Credits are additive if | Q3 2026 reading |
|---|---|---|
| NDR above $50k ARR | Stabilises or rises while credit adoption grows | 114%, down from 115% |
| AI ARR definition | Top-up and stand-alone revenue is broken out from bundled credits | Not broken out |
| Seat commentary | Paid seat growth is quantified alongside AI ARR | Directional only, no figure |
| Net adds above $100k | Accelerate as credit spend lifts account size | Slowed from 175 to 108 |
| Where monday.com wins | Over 30% of $50k accounts pay for credits | Reported, and a real base |
The last row concedes the strongest point in monday.com's favour. Getting roughly 1 in 3 large accounts to pay for a new meter within six months is hard. Most vendors that tried bundled credits, including the cases in HubSpot's Breeze credits repricing, struggled to get buyers past the included allowance at all.
The Q4 guide, computed from the two releases
The full-year guide looks like a raise. In August it was $1,466 million to $1,474 million. In November it was $1,473 million to $1,475 million. The midpoint moved up by $4 million.
The Q3 beat was bigger than that. The August guide for Q3 was $368 million to $369 million at the midpoint of $369 million, and Q3 came in at $377.5 million. That is $8.5 million above the midpoint.
So the implied fourth quarter came down. From the August guide, the implied Q4 midpoint was $1,470 million less $715.9 million of first-half revenue less $369 million, which is about $385.1 million. From the November guide it is $1,474 million less $1,093.4 million, or about $380.6 million. The implied Q4 fell by about $4.5 million.
The company's own Q4 range of $380 million to $382 million says the same thing. Glazer described it plainly on the call: "we're going down from 19% into 14%". He added that stabilisation was "not going to be a quick fix", likely running through 2027.
Part of that is cost and focus, not pricing. monday.com expects headcount to be about 20% lower by the end of 2026, and George said the disruption "lingered candidly a little longer than we thought". A company cutting a fifth of its staff while changing its pricing model has made the attribution problem harder for itself.
Where this reading of monday.com is weakest
My conclusion is that credits have not yet shown up as added revenue. Three facts cut against that conclusion, and each is strong enough that you should hold it loosely.
The retention lag and the price roll-off cut both ways
NDR is a trailing four-quarter average, so a credit model launched in May barely touches a September figure. Glazer also said in August that NDR would face pressure from lapping price increases taken in 2024 and 2025. A one-off seat price rise inflates retention for a year and then falls out of the comparison. Some or all of the decline may be that roll-off, not weak expansion. If so, credits could be working and the average simply has not caught up. I would not call the model a failure on retention data before the Q2 2027 report, when a full year of credit selling sits inside the average.
The SMB churn is chosen, and AI ARR is small on purpose
Management said the higher SMB churn was "by design", as it focuses on larger accounts. A deliberate exit from small customers lowers blended NDR while raising account quality. The $50,000 cohort still runs at 114%, which is healthy by most SaaS standards. On that view the headline 108% is a mix effect.
The AI ARR number is also small because credits are bundled into plans. A company that wanted a big AI headline could price credits separately and sell them hard. monday.com chose a bundle, which slows the visible AI line and may protect retention. The competing argument, made in the hybrid pricing NRR gap analysis, is that hybrid models take several renewal cycles to show their effect. That argument has real support, and this post does not refute it.
Frequently asked questions
What is monday.com's net dollar retention in Q3 2026?
monday.com reported overall net dollar retention of 108% in Q3 2026, down from 109% in Q2, 110% in Q1 and 111% in Q3 2025. Customers above $50,000 ARR were at 114%, and those above $100,000 at 113%. CFO Eliran Glazer said on the 9 November call that he expects overall NDR to fall another point to 107%.
How much AI ARR does monday.com have?
monday.com reported $36 million of AI ARR at the end of Q3 2026. The figure includes stand-alone AI products, AI credit packages and the share of subscription plans attributed to bundled AI credits. That last part means some AI ARR is reallocated plan revenue. The CFO said AI ARR is "not incremental yet" and remains part of overall ARR.
What is monday.com's seats plus credits pricing model?
Announced with Q1 2026 results on 11 May 2026, the model charges per seat for people and adds AI credits for work done by agents, the vibe app builder and other AI features. Plans include an allotment of credits, and customers who use their allotment can top up. Management says this gives accounts a second way to expand beyond adding seats.
Did AI credits increase monday.com revenue growth?
Not visibly yet. AI was 17% of net new ARR in Q2 2026 and reached $36 million of AI ARR in Q3. Over the same period, revenue growth slowed from 22% to 19% and Q4 is guided at about 14%. Management attributes the slowdown to SMB churn, longer sales cycles and restructuring rather than pricing, and calls credits generally additive.
Why did monday.com stock fall after Q3 2026 earnings?
Shares fell after the 9 November release despite a revenue beat. The release showed NDR down to 108% with an outlook of 107%, net customer additions slowing across every cohort, and Q4 revenue guided to about 14% growth. The full-year guide rose $4 million at the midpoint, less than the $8.5 million Q3 beat, implying a lower fourth quarter.
Is net dollar retention falling across SaaS companies with AI pricing?
It is falling at several companies that moved to hybrid seat and credit pricing, but the cause differs. At monday.com, management points to deliberate SMB churn and earlier price increases rolling out of the trailing average. Because NDR is usually a four-quarter average, a pricing change takes a year to show up fully, so early declines are not proof the model failed.
What to check before monday.com's Q4 report
If you run a SaaS business considering a seat and credit model, monday.com's next release is the most useful free data you will get this quarter. Read three lines in it, in this order.
First, look for any split of AI ARR between bundled credits and top-ups or stand-alone products. Only the top-up portion is evidence of expansion. If the release still blends them, treat the AI ARR figure as marketing rather than measurement. Second, check NDR for the $50,000 cohort. If it holds at 114% or rises while credit adoption grows past 30%, the additive case gets real support. If it falls again, the meter is replacing growth, not adding to it.
Then apply the same test to your own model before you copy it. Write down, now, what share of your planned AI revenue is allocated from plan price. If you cannot answer, you will be reporting a number like $36 million to your board without knowing whether any of it is new. The broader disclosure problem is ranked in how AI revenue metrics compare across filers.
Related on pricing and retention
For the same credit question at a design tool, read Figma's net dollar retention after AI credits.
References
- monday.com, Third Quarter 2026 Results, Form 6-K exhibit 99.1, 9 November 2026. Used for Q3 revenue, NDR, customer counts, AI ARR and its definition, and guidance. Retrieved 11 November 2026.
- monday.com, Second Quarter 2026 Results, 10 August 2026. Used for the 17% of net new ARR figure, Q2 NDR, customer counts and the August guidance.
- monday.com, First Quarter 2026 Results, Form 6-K exhibit 99.1, 11 May 2026. Used for Q1 NDR, customer counts and the seats plus credits launch.
- monday.com, Third Quarter 2025 Results, Form 6-K exhibit 99.1, 10 November 2025. Used for the year-earlier NDR baseline of 111%.
- The Motley Fool, monday.com Q3 2026 earnings call transcript, 9 November 2026. Used for all Q3 management quotes.
- The Motley Fool, monday.com Q2 2026 earnings call transcript, August 2026. Used for Q2 management quotes on top-ups, seats and the NDR outlook.
Weakest point in the source base: management quotes come from third-party call transcripts that contain visible transcription errors, and monday.com does not disclose total ARR, paid seats or the bundled share of AI ARR. Every calculation marked as Zan Digital's uses only figures from the 6-K releases.
Related reading