Cloudflare Inc · Moat analysis

Report previewNET.USReport date Last updated
← Back to reports
Report contents
01

Industry landscape

Where the company competes: one network, four separate fights

Cloudflare sells subscriptions to a single global network. A customer points its traffic at Cloudflare and buys, in bundles or one piece at a time, DDoS mitigation, a web application firewall, bot management, DNS, content delivery, zero trust access, and developer services such as Workers and R2. The company reports one operating segment and does not publish revenue by product, so any read on mix is an inference from management commentary. What it does disclose is geography and customer size: the United States supplied roughly half of 2024 revenue, EMEA the next largest share, and no single customer was material, the company said in its 2024 10-K.

Growth has slowed. Revenue rose 54% in 2022, 33% in 2023 and 29% in 2024, to $1.67 billion, and customers spending more than $100,000 a year numbered about 3,500 at the end of 2024. Deceleration at that scale is ordinary. The competitive question is where the incremental dollar now comes from, and which rival is positioned to take it.

The arena is not "cloud" in the abstract. Cloudflare fights in four adjacent submarkets that share a network, a sales force and a free tier, but not a set of opponents. Content delivery and authoritative DNS is the franchise business, sold on latency and uptime. Application security, meaning DDoS, WAF and bot defense, is sold to the same buyer and is where pricing is most visible. Zero trust and secure access sit in the security budget alongside firewall incumbents. Edge compute and storage put Cloudflare against the largest cloud providers on a product that is still early.

SubmarketMain rivalsShare evidence
Content delivery and DNSAkamai, Amazon CloudFront, FastlyCloudflare sits on about 80% of sites using a known CDN, W3Techs estimates; Akamai still earns more
DDoS and web application securityAkamai, Imperva, F5, AWSCloudflare is the most widely deployed by site count; enterprise revenue share is not public
Zero trust and SASEZscaler, Palo Alto Networks, Netskope, AkamaiCloudflare holds a single-digit revenue share, an estimate
Edge compute and storageAWS, Google Cloud, Fastly, VercelNo reliable share data; Cloudflare is an early challenger

Concentration differs by row. Adoption of Cloudflare's delivery product is close to a monopoly by site count, yet the revenue is not: Akamai reported about $4.0 billion of 2024 revenue growing roughly 5%, against Cloudflare's $1.67 billion growing 29%. Akamai earns more from fewer, larger delivery and security contracts. In zero trust, five or more credible vendors sell into the same budget, so no one sets price.

The buying unit is global. Anycast routing means a customer in Frankfurt is served by the same architecture as one in São Paulo, and the network spans more than 330 cities. That is why scale pays here in a way it does not in regional colocation: every new city lowers latency and absorbs attack traffic for the whole customer base at once. Data residency rules and the China partnership fragment the product, not the competitive set.

The barrier test is uneven. Building a comparable network takes years and sustained engineering, which shuts out startups; Cloudflare's early lead in the free tier and developer adoption reinforces that. It does not shut out the bundlers. Amazon, Microsoft and Google already own larger footprints and can give CDN and DDoS away against cloud commitments, while Palo Alto Networks, Zscaler and Akamai fold security modules into contracts that already sit in the customer's budget. Cloudflare's answer is to sell more products per account at a lower acquisition cost, which is why sales and marketing has run at roughly 40% of revenue. That cost ratio is the tell: the free tier generates leads, but closing enterprise security deals still requires an expensive direct sales force.

Market position: Contested. Cloudflare is the clear leader in one submarket, delivery by site count, and a subscale challenger in the two where the security budget actually sits. The Q3 2026 investor presentation, dated as of September 30, 2026, describes a reorganization around an "agentic AI-first operating model" that includes a workforce reduction, which suggests selling and support cost is the constraint management is attacking.

What would settle the question is whether Cloudflare converts adoption into price. The company does not disclose average revenue per customer or split growth between price and volume, so the evidence does not yet establish whether it is winning security budget on merit or discounting into it. The contrast with Akamai's revenue per contract is the number to watch.

02

Competitive advantage

Does the company have a moat? Moderate, and it rests on scale

Moat rating: Moderate. Cloudflare has built a real franchise in internet infrastructure and has grown revenue about 35% a year, from $0.66 billion in fiscal 2021 to $2.17 billion in fiscal 2025. It has also reported a net loss in every year since its 2019 listing. Those two facts sit awkwardly together, and the awkwardness is the assessment: a company can win share quickly without having shown that entry is blocked.

The two tests that settle the question are persistence of concentration and persistence of excess returns. Cloudflare passes the first only in part and fails the second outright. Where a leader holds share for years and earns above its cost of capital, a barrier is doing the work. Where the leader holds share and still reports negative returns, the share is more likely the product of spending and execution than of protection.

Market position: first by adoption, subscale where the budget sits

On the measure most favorable to it, Cloudflare is dominant. W3Techs estimates it sits behind roughly 80% of websites that use a known CDN, a share no rival approaches on that count. On revenue the ranking changes: Akamai reported about $4.0 billion of 2024 revenue, more than twice Cloudflare's, on growth near 5%. In zero trust and secure access, the submarket that absorbs enterprise security budgets, Cloudflare's revenue share is in the single digits, an estimate built by comparing its disclosed revenue with the size of that market.

Market position: Moderate. A company that ranks first in one submarket and is a challenger in the ones carrying the larger budget does not hold a broad structural position. Cloudflare reports a single segment and does not disclose revenue by product, so the split among delivery, security and compute is not observable, and the rankings above rest on third-party adoption counts and disclosed totals rather than on a company-reported share.

Market position trend: five years of gains, which is not stability

Revenue compounded at roughly 35% a year from fiscal 2021 through fiscal 2025, and growth reaccelerated after that. Cloudflare reported $0.64 billion in the first quarter of 2026 and $0.70 billion in the second, up about 33% and 37% from $0.48 billion and $0.51 billion a year earlier.

Winning share at that pace against Akamai, Amazon and the security incumbents shows the product is competitive. It does not show the position is protected, and the framework reads fast gains the other way. A rule of thumb in this framework: a share that moves less than 2 percentage points over five years points to strong barriers, while a move of more than 5 points suggests there are none. Cloudflare's share is moving quickly in its own favor, which says something about execution and nothing yet about entry barriers. Market position trend: Moderate.

Profitability: no year yet clears the cost of capital
USD bn unless statedFY2021FY2022FY2023FY2024FY2025H1 2026
Revenue0.660.971.301.672.171.34
Gross margin77.6%76.1%76.3%77.3%74.5%71.6%
Operating margin-19.7%-20.6%-13.8%-9.0%-9.7%-20.1%
Net margin-39.7%-19.8%-14.2%-4.7%-4.7%-14.2%
Operating cash flow0.070.120.250.380.600.28
Free cash flow-0.04-0.040.120.200.320.14
Return on equity-31.9%-26.9%-26.5%-8.7%-8.2%n.m.

There is no ten-year record to average: Cloudflare listed in September 2019. Over the five fiscal years above, return on equity averaged about negative 20%, and the best single year, fiscal 2025, was still negative 8.2%. Net margin improved from negative 39.7% in 2021 to negative 4.7% in both 2024 and 2025, then broke. The second quarter of 2026 produced an operating loss of $0.21 billion on $0.70 billion of revenue, as large as the entire loss of fiscal 2025. Whether that quarter carries a one-time charge or marks a permanently higher cost base is the most important open question in this section, and the reported figures do not separate the two.

The cash numbers are better and still fall short. Operating cash flow rose from $0.07 billion in 2021 to $0.60 billion in 2025; free cash flow went from negative $0.04 billion to $0.32 billion. Against $6.04 billion of total assets at the end of 2025, that is a return of about 5%, below the 15% the framework wants before it calls an advantage proven. The ratio understates the return on operating capital because total assets include goodwill and marketable securities, so treat it as a rough gauge rather than a precise figure. The $0.70 billion gap between the net loss and operating cash flow is non-cash charges, stock-based compensation among them, and working capital. That compensation rewards employees rather than suppliers, but shareholders still bear it. Profitability: Weak.

Margins get a separate reading. A gross margin between 71% and 78% is high in absolute terms and implies the product carries real value to customers. The problem is what happens below the gross line: sales and engineering costs consume the whole spread and then some. On revenue scale, Cloudflare remains roughly half of Akamai and a fraction of the cloud providers it now competes with. Margins: Moderate.

Overall rating: Moderate, with the profitability test unanswered

Market position and its trend both rate Moderate. Profitability rates Weak, and the framework's instruction is to weight profitability heavily: a company that has not earned its cost of capital has not demonstrated a moat, whatever its gross margin says. Margins rate Moderate.

That leaves Moderate overall. The case for anything higher rests on the network-scale argument in the next section plus the steady improvement in operating margin from negative 20% in 2022 to negative 10% in 2025. The case for anything lower rests on the failure of every return measure and the second-quarter 2026 reversal.

What would settle it: operating margin returning to about negative 10% while gross margin holds above 74%, and free cash flow climbing toward 8% to 10% of total assets. If gross margin keeps sliding and the operating loss stays at the second-quarter level, the scale claim weakens materially.

Where the advantage comes from and how long it can last: network scale and developer habit

Two mechanisms support the Moderate rating, and they reinforce each other. One is the cost structure of a shared network. The other is the stickiness of the layer where customers first touch Cloudflare. Both are real, and both are narrower than the company's growth story implies.

Economies of scale: real, relative, and thinner in compute

Cloudflare runs a single anycast network, so a request from Frankfurt and one from São Paulo are served by the same architecture from the nearest of more than 330 cities. The fixed costs are the points of presence, the servers inside them, the transit and peering contracts, and the engineering needed to keep all of it running safely. Those costs scale with capacity, not with the number of customers.

The consequence is straightforward. Opening a city raises performance for every existing customer at once, so a fixed outlay is spread across the whole base rather than a segment of it. Adding a customer to an existing city adds revenue at a low incremental cost. That is why gross margin sits in the mid-70s even while the company is spending heavily to grow, and it is the strongest structural fact about the business.

Two limits keep this from rating higher. The first is relativity: Akamai and the three large cloud providers run networks of comparable or greater reach, and the cloud providers spread network cost across compute and storage revenue that Cloudflare does not have. Scale advantages accrue to the leader in a specific market, and Cloudflare is the leader only in delivery by site count.

The second limit is the direction of mix. Gross margin fell about 3 percentage points in 2025 and another 3 points in the first half of 2026. A shift toward storage and compute products, which sell at lower gross margin than CDN and DDoS, may explain it. Cloudflare does not disclose product-level margins, so that is an inference rather than a reported fact. If it is correct, the scale advantage gets thinner as the company grows, because incremental revenue arrives at a lower spread and the fixed-cost leverage works on a smaller base. Economies of scale: Moderate.

Customer captivity: sticky at the DNS layer, thinner in security

The first product most customers buy is also the hardest to leave. Authoritative DNS is configured once, at the registrar, and changing it means re-pointing nameservers, reissuing certificates, re-testing firewall rules and accepting a window in which a mistake takes the site down. That is a genuine switching cost measured in engineer hours and outage risk, not in contract terms, and it explains why a free tier converts into paid business.

A second layer sits underneath: code written for Workers runs on Cloudflare's runtime and does not port cleanly elsewhere. Once a customer puts application logic on the platform, the cost of leaving rises with every release. The developer free tier compounds this, because engineers learn the product on their own projects and later introduce it at work, which shortens the sales cycle and lowers customer acquisition cost.

The captivity weakens at the security layer, where the budget actually is. A chief information security officer compares Cloudflare with vendors already on the approved list, and web application firewall and bot rules can be rebuilt on a rival platform with effort but without architectural change. When Palo Alto Networks or Zscaler folds a competing module into a contract the customer has already signed, the Cloudflare alternative is not competing against zero; it is competing against a line item the buyer has already paid for. Cloudflare does not publish gross or net revenue retention, so the strength of this captivity cannot be measured directly, and the evidence remains indirect. Customer captivity: Moderate.

What is not a moat here: licenses, capital and brand

It is worth being explicit about the sources that do not apply. Cloudflare holds no operating license that a rival cannot obtain, and regulation works against it as often as for it, since data residency rules force the network to be replicated country by country. Patents do not block entry into CDN or DDoS mitigation. The balance sheet, with $1.66 billion of cash against $3.35 billion of debt at June 30, 2026, gives the company room to invest, but access to capital is a resource, not a barrier. The brand is strong among developers and weakens as a defense the moment a competitor discounts.

Durability: the bundle is the substitute to watch

The threat to this franchise is not another CDN. It is the cloud contract. Amazon, Microsoft and Google can include content delivery and DDoS absorption inside a committed-spend agreement at close to zero incremental price, which caps what Cloudflare can charge for the products that sit closest to the network. Cloudflare's answer is to sell more products per account and to price below the bundle, which is a strategy that depends on continued product breadth rather than on a protected price.

A second, more speculative substitute is a change in the shape of traffic. If a growing share of requests comes from machine-to-machine and agent traffic rather than page views, the value per request of a delivery network may fall, and with it the pricing that funds the shared infrastructure. The evidence so far only shows that request volume keeps growing; it does not establish how the mix will shift.

Weighing it together: the delivery and DNS core looks defensible for several years, protected by the network cost structure and by the nuisance of moving a nameserver. The security and compute expansion, where the growth has to come from, is where the moat is thinnest, because the opponent there is a bundle rather than a product. That is why the rating is Moderate: the strongest barrier protects the smaller revenue pool, and the largest opportunity is the most contested.

Generated by DeepVal AI. Data and models may contain errors or be outdated. For research only; not personalized investment advice.

Continue the comparison

Continue reading the full report

  1. Comparison and summaryHow Cloudflare Inc compares with its closest rivals, and our moat rating

Free to sign up · 20 research points · unlocking this report uses 10

20 points cover 2 full reports. No card needed.

Research reflects its generation date. Preview the first two modules.

More on Cloudflare Inc