A 20% Workforce Cut on the Best Bookings Quarter Since 2023, and a Full-Year Raise That Is Only the Q1 Beat

Results vs. Consensus

Cloudflare reported the March quarter after the close on May 7, 2026, and announced a reduction of approximately 1,100 roles in the same release. The quarter itself was not the problem. Revenue of $639.755M cleared the $622M Street figure by $17.8M and the company's own February guide midpoint of $620.5M by $19.3M, the second consecutive quarter of clearing guidance by more than $19M. Non-GAAP operating income of $73.1M cleared the $70.0M to $71.0M guide, and non-GAAP EPS of $0.25 cleared the $0.23 both guided and expected.

Q1 2026 Scorecard

Year-over-Year Comparison

Sequential Comparison

The sequential table is where the ambivalence lives. Q1 is seasonally the weakest quarter for this business and a step down from Q4 on margin is normal, but the size of the step is not: 320bps of operating margin and 210bps of gross margin in one quarter, with the gross-margin piece explicitly guided to continue.

Revenue

Revenue of $639.755M grew 33.5% year over year and 4.1% sequentially. The composition is what makes the number credible rather than lucky. Growth was almost perfectly uniform across geographies (United States +34.5%, Asia Pacific +34.4%, other regions +34.2%, Europe/Middle East/Africa +31.2%), which is the signature of a demand-led quarter rather than one large deal or one strong territory. The large-customer cohort did the heavy lifting: customers paying more than $100,000 annually reached 4,416, up 25.2%, and their share of revenue rose to 72% from 69%. Management disclosed that deals over $1 million grew 73% year over year and that the company added as many customers spending above $5 million annually in Q1 as it did in all of 2025.

The one number in the revenue block that gave the Street pause was dollar-based net retention at 118%, up seven points year over year but down two points sequentially. Management pre-empted the question by attributing the sequential move to mix rather than to expansion weakness, noting that growth in the quarter was driven by an unusually strong contribution from new customers, which does not enter the retention calculation. That explanation is arithmetically coherent and consistent with the disclosed new-customer bookings figure. It is also unfalsifiable from outside, which is why a second consecutive sequential decline in Q2 would matter considerably more than this one.

Margins

Non-GAAP gross margin of 72.8% is the lowest print in the company's recent history and the second consecutive sequential decline. Management gave two reasons, both structural rather than cyclical. The first is a reclassification effect: traffic from free customers, historically carried in sales and marketing as a customer-acquisition cost, is converting to paid traffic, and its network cost moves into cost of revenue with it. The second is the Workers developer platform, which carries a below-corporate-average gross margin and is the fastest-growing part of the business.

"The biggest move in gross margin this last quarter was free traffic moving to paid traffic and costs moving into cost of revenue. While that decreases gross margin, it is literally a wash from an overall P&L perspective."
— Thomas Seifert, CFO

The reclassification argument is correct as far as it goes and is visible in the numbers: non-GAAP sales and marketing fell to 36% of revenue from 38%, and total non-GAAP operating expense fell 400bps to 61.4%, so the cost did move rather than appear. The Workers argument is the one that does not wash. A structurally lower-margin product growing faster than the company average mechanically lowers blended gross margin for as long as it out-grows, and management said as much.

"While gross margin may continue to trend down in the near term from these dynamics, the scalability and efficiency of our network remain intact, and we expect our unit economic margin will continue to increase."
— Thomas Seifert, CFO

Assessment: The mix explanation is honest and the P&L-neutrality claim on the reclassification piece holds. But "gross margin may continue to trend down" from a company whose entire valuation case rests on software economics is a materially different disclosure than it would be from a company trading at ten times revenue, and it landed on the same afternoon as a request to be measured on a different metric.

EPS and Cash Flow

Non-GAAP EPS of $0.25 grew 56% against 33.5% revenue growth, which looks like operating leverage and mostly is not. Non-GAAP operating margin fell year over year. The gap between 33.5% revenue growth and 61% non-GAAP net income growth is carried by non-operating income, which rose to $40.6M from $16.5M, driven by interest income on a $4.16B cash and securities position that will not compound at that rate. GAAP net loss narrowed to $(22.9)M from $(38.5)M, and the GAAP-to-non-GAAP bridge remains large: $127.5M of stock-based compensation and related payroll taxes in a quarter with $639.8M of revenue, or 19.9% of revenue, which is high even by the standards of high-growth infrastructure software.

Free cash flow of $84.1M (13% of revenue) versus $52.9M (11%) is the cleanest number in the release and the one least discussed. Operating cash flow rose only 8.6% year over year, so the free-cash-flow improvement is chiefly a capital-intensity story: purchases of property and equipment fell to $65.2M from $85.9M. Network capex ran at 9% of revenue in Q1 against full-year guidance of 14% to 15%, which means the capex line is a headwind to free cash flow for the balance of the year, not a tailwind.

Segment Performance

Cloudflare reports a single operating segment. The two disclosed disaggregations that carry analytical content are revenue by billing region and revenue by customer type, both from the Q1 10-Q. Management narrates the business through a four-"Act" framework (application services, Zero Trust and SASE, the Workers developer platform, and AI content monetization) but does not report revenue against it, so the Acts are discussed in the commentary section below rather than tabled here.

Revenue by Region

United States

The domestic business grew 34.5% and held at 49% of revenue, unchanged from the prior-year quarter. That stability is the point: Cloudflare's US business is now large enough ($1.26B annualized on this quarter's run rate) that maintaining a growth rate above the corporate average is a meaningful result rather than a small-base artifact. The named wins management described in the quarter skew heavily US enterprise, including an $8M two-year Fortune 100 privacy-proxy expansion and a $5M three-year Fortune 500 aerospace and defense Zero Trust deal that displaced an incumbent whose browser-isolation product could not meet government compliance requirements.

Assessment: The competitive displacement pattern in these deals is more informative than the growth rate. Cloudflare is winning against first-generation Zero Trust vendors on compliance and against hyperscalers on latency, which are different arguments from the price-and-bundle competition that historically defined this market.

Europe, Middle East, and Africa

EMEA grew 31.2%, the only region below the corporate average, and held at 28% of revenue. The EMEA case study management described (a European insurer standardizing on Cloudflare for application services and full SASE under a five-year $5.1M contract, displacing six legacy vendors at signing with ten more underway) is the consolidation motion this region is supposed to deliver, and management noted the customer targets over $1.3M in annual savings from the displacement.

Assessment: A three-point growth gap versus the corporate average is not yet a trend, but EMEA is where the data-sovereignty argument should convert fastest given the regulatory backdrop, and it is currently the slowest region. Watch whether the sovereignty positioning management spent airtime on in Q&A shows up here before it shows up in the narrative.

Asia Pacific

APAC grew 34.4% and held at 15% of revenue. Management's APAC illustration was a two-year $8.7M expansion with a company whose own growth is being driven by AI-assisted software development, chosen over a competitive hyperscaler bid. This is the clearest single instance in the release of the second-order AI thesis: Cloudflare is not selling AI, it is carrying the traffic that AI-native companies generate.

Assessment: APAC growing in line with the US on a base one-third the size is the healthier read. The region is small enough that a single large expansion moves the growth rate, so treat one quarter of outperformance as noise and the two-year pattern as signal.

Other regions

The residual 8% of revenue grew 34.2%, in line with the corporate average and unchanged as a share of the mix. No commentary was offered and none is required at this size.

Assessment: Immaterial to the thesis. Its value is confirming that the growth was broad rather than concentrated.

Revenue by Customer Type

This is the most underdiscussed table in the filing. Channel-partner revenue grew 71.4% and now represents 30% of the total, up from 23%. Direct revenue grew 21.9%. Nearly half of the company's incremental revenue dollars in the quarter came through partners.

Assessment: Two readings, and they point in opposite directions. The favorable one is that a partner-led motion is what allows Cloudflare to protect quota-carrying capacity while cutting 20% of its workforce, and management explicitly said the channel is most effective on the consultative Act 2 products. The cautionary one is that a business shifting a rising share of revenue through intermediaries has less direct visibility into end-customer health and typically concedes margin, which is one plausible and entirely unaddressed contributor to the gross-margin decline. Neither the release nor the call quantified partner economics.

Key Operating Metrics

Key Topics & Management Commentary

Overall Management Tone: Two registers, imperfectly joined. On the business, management was the most confident it has sounded in several quarters, presenting the AI re-platforming as a demand event already visible in bookings rather than a future opportunity. On the workforce reduction, the founder-CEO carried the call personally and unusually candidly, opening and closing on the human cost rather than the savings, and repeatedly refusing the cost-cutting frame. What management did not do, in either register, was connect the two: the reorganization was defended on philosophy and the guidance was defended on prudence, and no one supplied the arithmetic that links a 1,100-person reduction to the $39.5M of operating income added to the year.

1. The Reduction in Force, and the Framing Management Chose

Cloudflare announced the removal of approximately 1,100 roles, roughly 20% of a workforce that ended the quarter at about 5,500. The filings put the cost at $140M to $150M, comprising $105M to $110M of cash for notice periods, severance and benefits and $35M to $40M of non-cash charges for accelerated vesting of share-based awards. The majority lands in Q2 and execution is expected to be substantially complete by the end of Q3.

Management refused the cost-reduction framing at every opportunity, positioning the action as a re-architecture of how work is done rather than a response to any financial pressure.

"This is not a cost-cutting exercise or an assessment of the individuals' performance. It is about defining how a world-class, high-growth company operates and creates value in the agentic AI era."
— Matthew Prince, co-founder & CEO

The supporting evidence offered was internal rather than financial: AI usage inside the company up more than 600% in three months, 97% of engineering using AI coding tools built on the same Workers platform sold to customers, and every engineering contribution to production code bases now reviewed by autonomous agents. The internal harness has a name, Cloudflare OS, and management said it is being demonstrated to customer CIOs.

Assessment: The framing is coherent and probably sincere, and the decision to lead with the human cost rather than the savings is the opposite of what a company hiding a demand problem would do. It is also the framing that makes the action hardest to underwrite. A cost-cutting exercise has a quantifiable payback; a philosophical re-architecture does not, and management gave the market no run-rate savings number to hold it to.

2. The Guidance Arithmetic: A Raise That Is Exactly the Q1 Beat

This is the single most important number in the release and it was never discussed on the call. In February, Cloudflare guided full-year 2026 revenue to $2,785M to $2,795M, a midpoint of $2,790M. It now guides $2,805M to $2,813M, a midpoint of $2,809M. The raise is $19.0M. The Q1 beat against the company's own guide midpoint was $19.3M.

The CFO addressed conservatism when asked, but about the restructuring risk rather than about the tailwind.

"We have been careful to reflect whatever residual risk remains in the guidance for the remainder of the year. As usual, we have tried to be thoughtful and prudent in how we think about what is in front of us."
— Thomas Seifert, CFO

Assessment: There are two clean readings and the stock traded on the less charitable one. The charitable reading is that management deliberately built a buffer for execution risk during a 1,100-person reorganization, which is exactly what a prudent CFO should do and which sets up beats in Q3 and Q4. The uncharitable reading is that the AI demand narrative is real but not yet large enough to change the revenue trajectory, and that a company growing 33.5% guiding the balance of the year to roughly 28% is telling you where it actually is. Both readings are available; only time separates them, and at 24x revenue the market is not obliged to wait.

3. The Operating-Income Bridge, and Where It Sits

Full-year non-GAAP operating income guidance rose from $378M–$382M to $418M–$421M, a $39.5M increase at the midpoint on a $19.0M revenue increase. That is the real raise, and its timing is the problem. Q1 delivered $73.1M at an 11.4% margin. Q2 is guided to $90M–$91M, a 13.6% margin. The remaining $255.9M against $1,504.7M of implied second-half revenue is a 17.0% operating margin, against 14.0% delivered for all of 2025.

Revenue and operating income for Q2, H2 and FY2026 are guidance midpoints; the H2 line is the residual after subtracting reported Q1 and guided Q2 from the full-year midpoint.

Assessment: A 340bps step-up in operating margin from Q2 to the second half, contingent on a reorganization that management says will not be substantially complete until the end of Q3, is the definition of an unearned raise. It is not unreasonable (removing 20% of a cost base should produce roughly this) but it is entirely prospective, and the company is asking to be paid today for savings it will not book until October.

4. Free Cash Flow "Unchanged" Despite the Cash Severance

The most quietly aggressive claim on the call was that full-year free-cash-flow expectations are unchanged despite $105M to $110M of cash severance landing mostly in Q2.

"These actions will result in severance and other restructuring charges of $140 million to $150 million for full-year 2026, approximately $40 million of which is noncash, with the majority concentrated in the second quarter. Our expectations for free cash flow for 2026, however, remain unchanged, with approximately 25% to 30% of full-year cash generation in the second and third quarters."
— Thomas Seifert, CFO

Absorbing roughly $107M of unplanned cash outflow without changing the full-year cash target requires either that offsetting in-year payroll savings roughly match the severance, or that the prior expectation carried that much cushion. Neither was stated, and no full-year free-cash-flow dollar figure has ever been published for 2026, so the claim cannot be checked against anything.

Assessment: The Q2-and-Q3 concentration comment is the tell. Guiding to 25% to 30% of full-year cash generation across two quarters means roughly 70% to 75% falls in Q1 and Q4, and Q1 has already delivered $84.1M. That shape is consistent with severance consuming the middle of the year, which is a legitimate answer, but it is an answer the market has to reverse-engineer.

5. Gross Margin and the Request for a New Scorecard

Asked directly how to think about gross margin as the lower-margin Acts scale, the CFO conceded the structural point and proposed a different measure.

"The margin structure is different across the various Acts, with developer products being the weakest on gross margin. Despite that, all products are equal when we look beyond gross margin to unit economic value. We will get you ready on Investor Day for thinking about operating margin as a better measure of product competitiveness than gross margin."
— Thomas Seifert, CFO

The CEO added the demand-side rationale: the free customer base has turned out to be largely developers, the developer products are mostly paid, and so historic free traffic funded out of the marketing line is converting into paid traffic carried in cost of revenue. On his framing the gross-margin optics understate what is happening.

Assessment: The economics of the argument are defensible. The timing is not. Asking investors to switch from gross margin to an undefined "unit economic margin", to be explained at the June 9 Investor Day, in the same quarter gross margin fell 210bps and was guided lower, is how a genuine mix story gets mistaken for goalpost-moving. Management should publish the metric before it needs it.

6. Rule of 40, and the Number Management Wants to Be Judged On

The CFO volunteered a composite framework in place of the margin detail he declined to give.

"With the guidance in place, we are getting north of 46% from a Rule of 40 perspective, and we think we have visibility to reaching north of 50% next year."
— Thomas Seifert, CFO

The current-quarter arithmetic supports the first half of that: 33.5% revenue growth plus a 13% free-cash-flow margin is 46.5%. The forward claim of north of 50% in 2027 is the first time management has put a number on what the agentic operating model is supposed to be worth, and it is the only quantified benefit of the restructuring anywhere in the disclosure.

Assessment: This is the most useful sentence on the call and it deserved a follow-up it did not get. North of 50% in 2027 on decelerating revenue growth implies a free-cash-flow margin in the low-to-mid twenties, roughly double the current level. That is a specific, checkable commitment, and it is the single thing an investor should hold this management team to over the next four quarters.

7. Bookings and the Large-Customer Cohort

The leading indicators were uniformly strong and were the part of the release nobody disputed. Sales productivity rose year over year for a ninth consecutive quarter. New-customer bookings grew at the highest rate since 2023. New pipeline generation grew sequentially at the fastest pace in five years and exceeded plan by more than any first quarter since 2021. Gross retention hit a four-year high.

"Deals over $1 million were up 73% year-over-year, the fastest growth rate since 2024. We added a record number of our largest customers in the quarter—those spending more than $5 million with us annually. In fact, we added as many $5 million-plus customers in Q1 as we did in all of last year."
— Matthew Prince, co-founder & CEO

Assessment: If the restructuring had not been announced on the same afternoon, this paragraph is the quarter. A record cohort of $5M-plus customers, $1M-plus deals up 73%, and four-year-high gross retention is not the profile of a business with a demand problem, and it is the strongest available evidence that the reorganization is being run from strength rather than necessity.

8. The Workers Developer Platform

The developer platform added a million developers in a single quarter against 1.5 million in all of 2025, which is the sharpest inflection in any disclosed metric.

"In fact, developers on our platform increased to more than 5.5 million at the end of the first quarter—an increase of 1 million developers in a single quarter, as compared to an increase of 1.5 million in all of 2025."
— Thomas Seifert, CFO

The CEO supplied a usage datapoint on the newer Dynamic Workers primitive, positioned against containers for agentic workloads.

"One large AI studio, in just the last 15 days, went from essentially zero Dynamic Workers to over 1 million Dynamic Workers running across our platform."
— Matthew Prince, co-founder & CEO

Assessment: Developer count is the best forward indicator this company discloses and the worst monetization indicator, and both facts matter. The platform is winning mindshare at a rate that would be hard to reverse. It is also, by management's own statement, the weakest product line on gross margin, so every unit of this success dilutes the metric the market currently uses to value the company. That tension is the thesis in miniature.

9. Capital Intensity: Investing Behind Demand Rather Than Ahead of It

Pressed on whether Cloudflare would add capacity to chase inference workloads, the CEO drew a sharp distinction between his model and the hyperscalers'.

"Across hyperscalers, GPU utilization is often in the single digits. We are getting our GPU utilization to approach our CPU utilization—up in the 70% to 80% range. As we do that, we can keep servicing requests with the fleet we have and invest behind demand, rather than ahead of demand."
— Matthew Prince, co-founder & CEO

The financial expression of this is a capex line that ran at 9% of revenue in Q1 against 14% to 15% guided for the year, and a free-cash-flow margin that expanded 200bps year over year primarily because property and equipment purchases fell in absolute dollars.

Assessment: This is the most durable structural argument in the Cloudflare story and the one least reflected in the current debate. If the claim about utilization holds, Cloudflare participates in inference demand without the balance-sheet commitment that is currently compressing returns across the AI infrastructure complex. The claim is also unverifiable from outside and the capex guide implies the balance of the year is considerably heavier than Q1, so treat the 9% print as timing rather than a new run rate.

10. The Channel Motion and Sales Capacity Through the Reorganization

The commitment that most directly protects the revenue line through the restructuring is that quota-carrying sales capacity was exempted.

"While this action affects all teams at Cloudflare, Inc., the only real exception is our AE and quota-carrying capacity sitting in front of customers—we hardly touched that."
— Thomas Seifert, CFO

The mechanism described is a change in support ratios rather than a change in headcount budget: fewer support roles per account executive frees dollars to fund more quota-carrying capacity inside the same spend envelope. Management also said hiring in Q1 was majority sales, and that growth in net quota-carrying capacity is still expected to accelerate in 2026.

Assessment: Protecting the revenue-generating layer is the right call and the reason the revenue guide is defensible at all. The unhedged risk is the support layer itself: sales engineers, solution architects, customer success and deal desk are exactly the "support roles" being thinned, and they are load-bearing for the consultative Act 2 motion and the $1M-plus deals that grew 73%. A pure account-executive count can rise while capacity to close complex deals falls.

11. Act 4, Pay-Per-Crawl, and the Optionality Nobody Is Paying For

The CEO devoted the longest single answer of the call to the fourth Act, the attempt to build a monetization layer between AI crawlers and content owners.

"On Act 4, we think the business model of the Internet—historically advertising—is about to change dramatically over the next five years. It may not change to one thing but several. Because of how much of the Internet sits behind Cloudflare, Inc., we have a seat at the table in defining that."
— Matthew Prince, co-founder & CEO

The stated milestones are concrete for a business with no disclosed revenue: penetration in media has gone from low to dominant, media executives report striking better deals with AI companies because of the control tooling, and the 2026 objective is first revenue that can be passed back to the long tail of content creators. Management also flagged microtransactions on agent requests as the mechanism it is watching, and said the constraint is finding partners who can handle the transaction volume.

Assessment: Act 4 is real optionality with a genuine structural advantage behind it, and it is correctly valued by the market at approximately zero today. It contributes nothing to the 2026 model. Its relevance to this quarter is what it says about capital allocation: a company cutting 20% of its workforce is simultaneously funding a speculative new business model, which is consistent with management's insistence that this was not a cost exercise.

Guidance & Outlook

Implied quarter-over-quarter ramp: the full-year midpoint of $2,809.0M less reported Q1 of $639.8M and guided Q2 of $664.5M leaves $1,504.7M for the second half, or 27.9% year-over-year growth against $1,176.5M in the second half of 2025. The guide therefore embeds a deceleration from 33.5% in Q1 to 29.7% in Q2 to roughly 28% across H2, with no re-acceleration from the AI demand management spent the call describing.

Street at: the Q2 revenue guide brackets consensus of approximately $665M at its top end and sits a fraction below at the midpoint. The full-year revenue raise is smaller than the Q1 beat that funded it. Only the operating-income and EPS lines were raised in substance, and both are second-half weighted.

Guidance style: consistently conservative on revenue and, on this evidence, newly aggressive on margin. Cloudflare cleared its own revenue guide midpoint by $25.5M in Q4 2025 and by $19.3M in Q1 2026, and the CFO's language about reflecting "residual risk" is the language of a team building buffer. The margin guide is the departure from pattern: management has never before asked the Street to underwrite a 340bps second-half operating-margin step-up.

Analyst Q&A Highlights

Why Restructure From a Position of Strength

The opening question of the call went straight to the contradiction, and it framed the entire session: a company posting its best bookings quarter in three years does not usually remove a fifth of its workforce on the same afternoon. Management's answer was explicitly not financial. The argument was that the roles the company needs have changed, that the productivity gains have concentrated in the people talking to customers and writing code, and that the support layer behind them is the part being re-architected. Management went further than required, forecasting that headcount would exceed 2026 levels again by 2027.

Q: "And then as a follow-up, regarding the announced restructuring, really in light of these strong Q1 results, it seems to be coming from a position of strength. Why now? How is it going to make Cloudflare, Inc. stronger? And, Thomas, have you embedded any conservatism in the guide for this action?"
— Matthew Hedberg, RBC Capital Markets

A: "I would guess that in 2027 we will have more employees than we did at any point in 2026, but the roles are changing dramatically, and you have to do something dramatic to make that shift. That is why this is the right time. We are the fittest we have ever been, but we are going to get even fitter to win the next chapter."
— Matthew Prince, co-founder & CEO

Assessment: The answer is unusually falsifiable, which counts in its favor. A company that intends to be larger in 2027 than in 2026 is not managing to a cost target, and that claim can be checked. What the answer does not do is supply any bridge from the reorganization to the $39.5M of operating income added to the year, which is the number investors actually needed explained.

Conservatism Embedded in the Guide

The follow-up half of the same exchange was the one that mattered to the model, and it received a shorter answer. The CFO confirmed that account-executive capacity was effectively exempted from the action and that residual execution risk had been reflected in the balance-of-year guide, without quantifying either the exemption or the buffer.

Q: "And, Thomas, have you embedded any conservatism in the guide for this action?"
— Matthew Hedberg, RBC Capital Markets

A: "On guidance and how this action is reflected, we have been thoughtful. While this action affects all teams at Cloudflare, Inc., the only real exception is our AE and quota-carrying capacity sitting in front of customers—we hardly touched that. We have been careful to reflect whatever residual risk remains in the guidance for the remainder of the year."
— Thomas Seifert, CFO

Assessment: This is the answer that reconciles the unchanged rest-of-year revenue outlook with the biggest-tailwind-in-company-history narrative, and it is the more charitable of the two readings available. It also means the revenue guide and the margin guide are being defended on opposite principles: revenue is conservative because of execution risk, while margin assumes the same execution succeeds. Both cannot be prudent.

The Internal Cost of Running on AI

A line of questioning that has become standard across software this year (whether the cost of AI coding tools offsets the productivity they deliver) drew the most differentiated answer of the call. The argument was vertical integration: most internal AI workloads run on Cloudflare's own network and GPU fleet, routed by the same AI Gateway product sold to customers, with only harder tasks escalated to frontier models at cost.

Q: "Matthew, one of the things we keep hearing about is how AI costs internally are really expensive, especially around R&D coding agents. How do you think about balancing AI coding adoption with the cost? What AI efficiencies are you looking to see across the organization, and how much of that is to offset some of those costs?"
— Adam Borg, Stifel

A: "As usage has gone up—600% in the last quarter—we have seen costs go up, but not nearly as much as some others. The least important reason is that most of the big AI labs are our customers, so we have good relationships and access to the best pricing and models. More importantly, we can often run those models on our own infrastructure."
— Matthew Prince, co-founder & CEO

Assessment: The most important disclosure buried in this answer is commercial rather than operational: management is demonstrating its internal harness to customer CIOs and signalling it may productize it. Cloudflare has a long record of shipping internal tools as products, so this is a credible pipeline item. It also means the same infrastructure that justifies the workforce reduction is being positioned as a revenue line, which raises the stakes on execution considerably.

Gross Margin Trajectory and the Net OpEx Effect

The most substantive analytical pushback of the call asked how to model gross margin as the lower-margin Acts scale and what the net operating-expense effect of the reorganization would be once complete. The answer conceded the gross-margin direction, reframed the metric, and did not address the operating-expense question at all.

Q: "Thomas, growth in different Acts has different impacts on gross margin, and you spoke about proactive optimization. As we get through those changes by the end of Q3, how should we think about the net impact to OpEx, and how should we think about gross margins as those other Acts continue to grow?"
— Saket Kalia, Barclays

A: "The margin structure is different across the various Acts, with developer products being the weakest on gross margin. Despite that, all products are equal when we look beyond gross margin to unit economic value. We will get you ready on Investor Day for thinking about operating margin as a better measure of product competitiveness than gross margin."
— Thomas Seifert, CFO

Assessment: The direct question on net operating-expense impact was asked and not answered, and no one returned to it. That omission is the reason the implied 17.0% second-half operating margin has to be reverse-engineered from a residual rather than read off a disclosure. In a quarter where the margin raise is the entire raise, declining to quantify the mechanism behind it is the most consequential non-answer on the call.

Willingness to Add Capacity for Inference Demand

Questioning turned to whether Cloudflare would follow peers in adding megawatts to host inference workloads. The CFO's answer was that nothing is being restricted, and the CEO reframed the question as a category error about business models, distinguishing an asset-leasing model that must build ahead of demand from a scheduling model that can build behind it.

Q: "Thomas, you mentioned Rule of 50 there potentially. Given demand behind inferencing, what is the team's willingness to go after more of this opportunity and drive more megawatts behind the network to host more of those inference use cases, like what you are seeing from some edge peers?"
— James Fish, Piper Sandler

A: "You see us leaning into this opportunity with all the force we have. That is why developer count went up by 1 million in the first quarter alone. We continue to optimize margin for these products, but we are not restricting growth—just the opposite. There is no restriction on leaning into this opportunity."
— Thomas Seifert, CFO

Assessment: The answer resolves a real bear point. Cloudflare is not rationing growth in the developer platform to protect reported gross margin, which would have been the value-destroying response to the mix pressure. The corollary the market has to accept is that blended gross margin keeps falling for as long as this is true.

Pool-of-Funds Renewals Reaching Maturity

The most useful commercial question of the call concerned the pool-of-funds contracting motion, now old enough that early adopters are reaching renewal. This is the first cohort large enough to test whether the model expands relationships or merely front-loads them.

Q: "Thomas, on the pool of funds and your serving year three in earnest of having this motion mature, any comments or observations as early pool-of-funds adopters come up for renewal? What trends are you seeing on renewal and expansion?"
— Gabriela Borges, Goldman Sachs

A: "We are now in our sixth quarter of pool of funds, and it has become a standard tool in our go-to-market motion. Teams are familiar with how and when to deploy it. From a renewal perspective, we had our highest-ever renewal rate last quarter, and that includes all pool-of-funds deals up for renewal."
— Thomas Seifert, CFO

Assessment: A record renewal rate inclusive of the first meaningful pool-of-funds cohort is the strongest single fact disclosed on the call, and it is the one that most directly supports the durability of the large-customer growth. It also pairs uncomfortably with dollar-based net retention falling two points sequentially: renewal strength and expansion strength are different things, and only the first was quantified.

Sales Capacity, Support Ratios, and the Channel Mix

The final exchange pressed on how quota-carrying capacity can keep accelerating through a 20% reduction, and on what is driving partner revenue to 30% of the total. The mechanism described was a change in the ratio of support staff to account executives, freeing budget for more quota carriers inside an unchanged spend envelope.

Q: "Thomas, you mentioned quota-carrying sales capacity continues to accelerate. Could you provide more color on expectations to continue to grow capacity relative to productivity? And for my follow-up, partners increased to 30% of revenue this quarter. What is driving this increase, and how much more channel mix would you expect going forward?"
— Shaul Eyal, TD Cowen

A: "When I said we are not touching quota-carrying AE sales capacity, what goes along with that is we see significant productivity gains in the support ratios for these AEs. The ratios are going to change significantly, which frees up dollars. Within the same spend envelope, you can deploy more quota-carrying AE capacity toward our market opportunity."
— Thomas Seifert, CFO

Assessment: This is the clearest statement of where the 1,100 roles actually come from, and it is the exchange that should worry a bull most. "Support ratios" in an enterprise software organization means sales engineering, solutions architecture and customer success, and those are precisely the functions that close the consultative Act 2 deals and the $1M-plus contracts that grew 73% this quarter. The bet is that agents can absorb that work. If it is wrong, the damage shows up in bookings two to three quarters out, well after the margin benefit has been booked.

What They're NOT Saying

1. The non-GAAP treatment of $105M–$110M of cash severance. The non-GAAP definition as filed in the Q1 10-Q excludes stock-based compensation, amortization of acquired intangibles, acquisition-related expenses, lease impairments and legal settlements. It does not exclude restructuring or severance. Under the definition as it currently stands, the cash severance sits inside non-GAAP operating income, in the same year that non-GAAP operating income guidance rose $39.5M. Whether the definition will be amended was neither asked nor volunteered.
2. Any run-rate savings figure. The cost of the plan is disclosed to the million. The benefit is disclosed nowhere. For an action management insists is not about cost, that asymmetry is convenient: it removes the only number against which the reorganization could later be judged a failure.
3. The net operating-expense impact, which was asked directly. The question was put plainly in Q&A and the answer addressed gross margin instead. The implied 17.0% second-half operating margin remains a residual investors must compute rather than a figure management stands behind.
4. A full-year free-cash-flow number. Expectations are described as "unchanged" against a target that has never been published. Absent a dollar figure, the claim that roughly $107M of cash severance changes nothing cannot be tested.
5. Which functions and geographies absorbed the cuts. "Across all functions and geographies" plus an exemption for quota-carrying account executives is the entire disclosure. The distribution matters enormously for revenue risk, and none of it was given.
6. Regretted attrition among the 4,400 who remain. No question was asked and no comment offered about voluntary departures following an announcement of this size. In a market where engineers with agentic-infrastructure experience are the scarcest resource in software, this is the largest unpriced risk in the quarter.
7. What "unit economic margin" is. The metric was invoked twice as the reason gross-margin erosion should be discounted. It has no published definition, no historical series and no disclosed level, and it was deferred to the June 9 Investor Day.
8. The retirement of the total paying-customer metric. The 10-Q discloses that the company has stopped reporting total paying customers because it "has become significantly less important to us over time in evaluating our business". That change was made in a filing, not on the call, in the same quarter dollar-based net retention declined sequentially.
9. Revenue by Act. The four-Act framework structured most of the prepared remarks and most of the Q&A, and carries a stated gross-margin ranking, but no Act has ever been sized. Investors are asked to accept a mix-driven margin explanation without the mix.

Market Reaction

• Pre-print setup: NET closed at $256.79 on May 7, up 30.3% year to date, up 111.0% over the trailing twelve months and up 18.7% over the trailing thirty days, at the very top of a 52-week closing range of $121.71 to $256.79. The S&P 500 was up 7.2% year to date over the same period. The stock entered the print priced for acceleration.
• After-hours move: shares fell roughly 18% to 19% in the May 7 extended session, with the decline beginning on the restructuring disclosure in the release rather than on the financial results.
• Next-day session: NET opened at $217.22 on May 8, a 15.4% downside gap, traded a $192.27 to $219.00 range, and closed at $196.13, down 23.6% or $60.66 on the day. Volume of 18.9M shares was 4.5 times the 30-day average of 4.2M.
• Index context: the S&P 500 rose 0.8% on the same session. The entire move was idiosyncratic.
• Value change: approximately $21.4B of market capitalization on 353.4M shares outstanding, against a company that raised full-year revenue, operating income and EPS guidance the same afternoon.

The shape of the move is more informative than its size. The stock fell 18% to 19% after hours and closed the next session down 23.6%, meaning the sell-off deepened by roughly five points during regular trading rather than stabilizing once the numbers were digested. Reactions that fade overnight are positioning; reactions that extend into the next session are re-rating. This was the latter.

Three things happened at once and only the second was actually a disappointment. The Q1 print was a clean beat. The Q2 revenue guide of $664M to $665M landed at or fractionally below a Street figure of approximately $665M, which for a stock trading at 32 times full-year guided revenue is functionally a miss, because that multiple requires the guide to move up, not sideways. And the restructuring converted a valuation question into an execution question overnight.

The deeper mechanism is a change in what the market thought it owned. Entering the print, NET was a straightforward long-duration AI beneficiary: agentic traffic grows, Cloudflare carries it, revenue compounds. Management then announced, in effect, that agentic AI is powerful enough to remove a fifth of its own workforce, and simultaneously left rest-of-year revenue unchanged. Read together those two statements say the AI effect is currently showing up in Cloudflare's cost base rather than its revenue line. A cost story and a growth story do not carry the same multiple, and the 23.6% move is roughly the distance between them.

Street Perspective

Debate: Is the Reduction in Force a Margin Step-Change or an Execution Risk?

Bull view: Removing 20% of the cost base while exempting quota-carrying sales capacity is the highest-return action available to a business with 33.5% revenue growth and only a 14.9% guided operating margin. The bull camp reads it as the moment Cloudflare stops being a growth-at-any-cost story and starts converting scale into profit, and points to the CFO's Rule of 40 comment as the first quantification of what that is worth.

Bear view: The skeptics note that no company has executed a reduction of this scale mid-year without disrupting a sales motion, that the "support roles" being removed are the sales engineers and customer-success staff who close complex enterprise deals, and that the damage from a botched reorganization appears in bookings two to three quarters after the savings are booked. On this view the margin benefit is front-loaded in the guide and the cost is back-loaded in reality.

Our take: The bear framing is closer to right on timing and the bull framing is closer to right on direction. The savings are real and roughly the right size; the risk is that they are recognized in the model before the revenue consequence is observable. The asymmetry favors waiting: an investor who buys after Q3 confirms both the margin step-up and clean bookings gives up some upside and avoids the entire distribution of outcomes where the reorganization goes wrong.

Debate: Is the AI Tailwind Actually in the Numbers?

Bull view: The bull case argues the tailwind is visible everywhere except the guide, which is deliberately conservative. Deals over $1M up 73%, record additions of $5M-plus customers, new-customer bookings at a three-year high, pipeline generation at a five-year high and a million new developers in one quarter are all leading indicators that convert to revenue with a lag. On this reading the guide is a floor and the second half is where the acceleration shows.

Bear view: The bear camp answers that a company genuinely seeing the biggest tailwind in its history would raise rest-of-year revenue by more than zero. The guide implies deceleration from 33.5% to roughly 28%, and management chose not to argue with that on the call. Bookings strength that never reaches the revenue guide is either not as durable as described or is being consumed by churn elsewhere.

Our take: The bull case has the better evidence and the bear case has the better discipline. Every leading indicator points up, and management's stated reason for conservatism (execution risk during the reorganization) is credible and internally consistent. But a guide is the only commitment that binds, and this one says 28%. Anyone underwriting acceleration is underwriting management's characterization over management's numbers, which is a defensible thing to do at 15 times revenue and a much harder thing to do at 24.

Debate: Does the Gross-Margin Decline Matter?

Bull view: The optimists treat the decline as pure mix and accounting. Free traffic converting to paid moves cost from marketing into cost of revenue with no P&L effect, and the developer platform carries a lower gross margin but a lower cost to book, so unit economics are unchanged or better. Total operating expense fell 400bps as a share of revenue, which is where the leverage was always going to show.

Bear view: The bears observe that gross margin has now fallen for two consecutive quarters, is guided lower again, and is 430bps below the prior-year quarter, and that a software company trading on software multiples cannot indefinitely explain away software-margin erosion. The request to be measured on an undefined metric, deferred five weeks to the June 9 Investor Day, is what a company does when the existing metric has stopped cooperating.

Our take: The mechanics favor the bulls and the disclosure favors the bears. The reclassification is genuinely P&L-neutral and the mix effect is genuinely a growth signal, so the operating-margin frame is the correct one. Management has nonetheless made this harder than it needed to be by asking for a new scorecard before publishing it. Publish "unit economic margin" with a back-series and this debate ends; leave it to the June 9 Investor Day and it compounds every quarter gross margin falls.

Model Update Needed

Valuation impact: at the May 8 close of $196.13 on 353.4M shares, NET carries a $69.3B market capitalization and, net of $4.16B of cash and securities against $3.27B of convertible notes, a $68.4B enterprise value. That is 24.4 times full-year guided revenue and 164 times full-year guided non-GAAP EPS, against 32.0 times and 215 times entering the print. On trailing-twelve-month free cash flow of $291.8M the stock trades at roughly 234 times enterprise value to free cash flow.

Fair value framework: we anchor on enterprise value to full-year guided revenue, the only forward metric management has committed to in dollars. At 25 times, the multiple the stock reached after the de-rating and roughly where high-growth infrastructure software with this growth profile has cleared, fair value is approximately $200 per share, implying about 2% upside from the $196.13 close. The bull case at 30 times, which requires the second-half margin step-up to land and revenue to re-accelerate rather than decelerate, supports roughly $240, or about 22% upside. The bear case at 18 times, which is where the multiple goes if the reorganization disrupts bookings, supports roughly $146, or about 26% downside. A distribution that skews slightly to the downside from a $200 anchor is the arithmetic behind the rating.

Thesis Scorecard Post-Earnings

This is our initiation on Cloudflare, so the pillars below are established here rather than carried forward, and this quarter's evidence is the first entry against them.

Overall: the operating thesis is stronger than we expected coming into the quarter and the valuation thesis is where the risk sits. Three of four bull pillars are confirmed on hard disclosure; the fourth, which is the one the raised guidance depends on, is entirely prospective. Both confirmed bear points were confirmed in the same afternoon.

Action: hold. The 23.6% de-rating removes the most acute valuation risk without making the stock cheap, and the two questions that decide the next twelve months (whether the second-half margin step-up lands, and whether bookings survive the reorganization) are both answered in the Q3 print. Adding ahead of that answer is paying for an outcome that becomes observable in October. We would move to Outperform on a Q3 report showing operating margin above 16% with $1M-plus deal growth intact, and to Underperform on a Q3 report where bookings decelerate while the margin benefit is still being described as pending.