Built by Claude FableAKAM | Thematic Exposure: 5/10
Deep dive pages: Overview | Business Model | Financial Trends | Thematic Exposure | Management Quality | Investor Sentiment | Concerns, Catalysts & Risks
Bottom line: Akamai is a co-leader in cloud web app and API protection (WAAP), which is a growing theme, and it has a real but small AI compute business that the Anthropic contract just made large. It does not have a verifiable >30% share in any meaningful segment. Delivery (27% of revenue) is shrinking and Akamai takes prices there. Cloud compute share is a rounding error next to the hyperscalers. The oligopoly hard gate is not met, so the score is capped at 5.
Segment mix and growth (Daloopa)
| Metric | Q3 24 | Q4 24 | Q1 25 | Q2 25 | Q3 25 | Q4 25 | Q1 26 | Q2 26 |
|---|---|---|---|---|---|---|---|---|
| Total revenue ($M) | 1,004.7 | 1,019.9 | 1,015.1 | 1,043.5 | 1,054.6 | 1,094.9 | 1,073.6 | 1,099.7 |
| Security ($M) | 518.7 | 534.6 | 530.7 | 551.9 | 568.4 | 592.4 | 589.8 | 604.4 |
| Security YoY | 14% | 14% | 8% | 11% | 10% | 11% | 11% | 10% |
| Delivery ($M) | n/a | n/a | 319.0 | 320.1 | 306.5 | 311.1 | 294.2 | 300.6 |
| Delivery + other cloud apps ($M) | 427.7 | 420.7 | 416.8 | 420.1 | 405.0 | 408.9 | 389.2 | 395.9 |
| Delivery + other YoY | n/a | n/a | n/a | n/a | n/a | n/a | -7% | -6% |
| Cloud Infrastructure Services ($M) | 58.3 | 64.7 | 67.6 | 71.5 | 81.2 | 93.7 | 94.6 | 99.3 |
| CIS YoY | 27% | 35% | 30% | 30% | 39% | 45% | 40% | 39% |
FY2025: total $4,208.2M, Security $2,243.4M (+10%), Delivery $1,256.7M, other cloud apps $394.1M, CIS $313.9M (+36%). Delivery YoY computed from Daloopa values: Q1 26 -7.8%, Q2 26 -6.1%.
Trajectory: Security is steady at 10% to 11% (decelerated from 14% in 2024; core WAF is slowing while API security grew >100% in 2025 to a >$100M run rate and API plus Zero Trust reached $90M in Q4 25, +36%, per the Q4 25 call). CIS accelerated from 27% to 39% to 45%, then held near 40% while the GPU capacity sold out. Delivery is in structural decline at -6% to -8%. Net result: mix is moving toward Security (55% of Q2 26) and Compute (9%), and away from Delivery (27%).
Required segment table
| Segment | % of Revenue (Q2 26) | Market Share | TAM | Theme Growth |
|---|---|---|---|---|
| Security (WAF/WAAP, DDoS, bot and API security, Guardicore segmentation, ZTNA) | 55.0% | Co-leader in cloud WAAP with Cloudflare and Imperva (Gartner Leaders). No revenue share figure published; estimate ~20% to 30% of cloud WAAP, unverified. Single digits in segmentation and Zero Trust | Cloud WAAP is several $B (QKS forecasts ~15% CAGR to 2030); broader app plus network security far larger | Theme +10% to 15%; AKAM +10% |
| Delivery (CDN, media, downloads) | 27.3% | Revenue share likely ~15% to 25% (one of 3 majors with Cloudflare and CloudFront). Usage share only 3.5% of websites (W3Techs-type trackers) | Cloud CDN ~$17B in 2025 per Reanin (low-quality estimate; revenue pools are deflating) | Traffic growing, revenue pool flat to down; AKAM -6% |
| Other cloud apps (edge apps, legacy) | 8.7% | Niche | n/a | AKAM -5% |
| Cloud Infrastructure Services (Linode, Akamai Inference Cloud) | 9.0% | Well under 1% of global IaaS (AWS, Azure, GCP control the majority) | IaaS several hundred $B; AI inference the fastest-growing part | Theme +20% to 30%; AKAM +39% |
Mix percentages from Q2 26 Daloopa values (Security $604.4M, Delivery $300.6M, other cloud apps $95.4M derived as $395.9M less Delivery, CIS $99.3M, total $1,099.7M). Share and TAM are third-party estimates and vendor reports; none give a vendor-level revenue share for WAAP, so these share figures are not verified.
Key competitors: Security has Cloudflare, Imperva (Thales), AWS WAF/Shield, F5, Fastly, and Zscaler/Illumio in segmentation and Zero Trust. Delivery has Cloudflare, Amazon CloudFront, Fastly, Google Cloud CDN, Edgio (gone), and in-house CDNs at hyperscalers and big streamers. Compute has AWS, Azure, GCP, CoreWeave, Nebius, DigitalOcean and Cloudflare Workers AI. Scale check: Cloudflare guides FY2026 revenue to $2,864M and printed $696M in Q2 26 (local NET workspace, press release), about 63% of Akamai's quarterly revenue while growing +36% YoY (FMP) vs Akamai's +5%. That gap is the core share-loss risk across both Delivery and Security.
Focus angle 1: edge compute capacity (physical asset)
- Footprint from management (Q1 26 and Q2 26 calls): ~4,300 locations in 700+ cities and 130 countries. Serverless functions run in all 4,300. Managed containers are live in "well over 100" cities. Full IaaS runs in "several dozen" cities, and "a couple dozen" of those have RTX Pro 6000-class (Blackwell) GPUs. The first ~20-city Inference Cloud GPU deployment sold out to beta customers before GA (Q4 25 call).
- No location count comes anywhere near this. CoreWeave, Nebius and IREN run tens of large campuses measured in hundreds of MW. Akamai's GPU capacity is a few dozen small metro deployments. On GPU count or MW it is far smaller than the neoclouds. On the number of sites where it can place compute and charge for it, it beats every neocloud and is comparable to Cloudflare (330+ cities; Fastly is smaller). Akamai does not disclose server count, MW or GPU count, so a per-site or per-$ comparison cannot be quantified. I have not invented one.
- Capex: 2026 includes ~$250M for Inference Cloud (Q4 25 call) plus ~$1.7B to pre-buy components for Anthropic, which takes FY2026 capex to about $3.5B (~78% of revenue). The Anthropic build needs ~$5.5B in total (8-K). The asset is under-appreciated as distribution. It is not under-appreciated as raw compute.
Focus angle 2: Anthropic deal (verified, 8-K dated 2026-09-24)
- Verified terms: $11.6B contractual commitment over 7 years for CPU workloads (the 8-K does not mention GPUs). Purchases can expand by up to $9B more (~$20B total). Anthropic gets a warrant on 7.7M shares (~5%) at $111.33: ~2% vests with the base commitment and ~1% per extra $3B. Akamai needs ~$5.5B of capex; the 8-K separately adds ~$1.7B to 2026 capex for component pre-purchases and does not say whether that counts toward the $5.5B. The 8-K says there is no impact to 2026 revenue guidance. Separately, Akamai has >$2.8B of other multi-year CIS commitments, including a ~$200M four-year AI Inference Cloud deal from Q4 25.
- Revenue timing (press coverage, not in the 8-K): $150M to $300M in 2027 starting in H2, rising to ~$1.7B annualized by end 2028. The 8-K confirms the commitment is subject to Akamai meeting delivery and service availability requirements, and Anthropic can terminate a plan on a material outage, subject to conditions. Treat it as firm-ish, not take-or-pay.
- Sizing: an average of $1.66B a year is ~5x FY25 CIS ($313.9M) and ~40% of FY25 total revenue. At run rate, CIS would rise from 9% to roughly 25% to 30% of revenue, and most of it would come from one customer. This explains most of the consensus revenue step-up from $4.2B (FY25) to $6.4B (FY28E).
- Does it make AKAM a "sneaky neocloud"? Partly. It is a neocloud-style deal: one AI lab, heavy capex up front, and an equity sweetener. But it is CPU capacity, not frontier GPU training, so it competes on distributed footprint and price, not on GPU scarcity. It raises thematic exposure. It does not create an oligopoly position, and the warrant is evidence that Akamai had to pay up to win it.
Focus angle 3: 10x web and API traffic scenario
| Segment | FY2025 base | Pricing mechanic | 5-yr bear | 5-yr base | 5-yr bull |
|---|---|---|---|---|---|
| Delivery | $1,256.7M | Committed traffic volume at a unit price that resets lower at renewal; multi-CDN moves bytes to the cheapest vendor | 0.6x ($0.75B): the FY2022 to FY2025 pace of -9%/yr continues | 0.8x ($1.01B): about -4.5%/yr, near the FY2025 rate of -4.7% | 1.0x ($1.26B): volume fully offsets deflation, which has not happened in any year of the series |
| Security | $2,243.4M | Sized to the protected estate and threat surface; agent traffic adds bot, API and WAF demand | +30% ($2.92B): ~5.4%/yr | +60% ($3.59B): ~9.9%/yr, the FY2025 rate | +100% ($4.49B): ~14.9%/yr, the FY2022 to FY2024 rate |
| CIS / edge compute | $313.9M | Capacity contracts plus pay-as-you-go usage | Contracted floor (~$14B signed) | Same | Upside from pay-as-you-go and the up to $9B Anthropic expansion |
Illustrative sensitivity, not a forecast. This is the single version used on the Overview, Thematics and Valuation pages: a five-year horizon from FY2025, calibrated to Akamai's own history. Akamai does not disclose traffic, so revenue per unit of traffic cannot be measured directly. The anchor is that Delivery revenue fell 25% from FY2022 to FY2025 (-9%/yr) while traffic grew, so even 10x traffic is unlikely to lift Delivery; the Security cases bracket its own recent growth rates of 5% to 15% a year. On net, 10x traffic helps Akamai mostly through Security, which is already 55% of revenue. That makes AKAM relevant to the theme, but it is not the main beneficiary. Cloudflare's developer and self-serve funnel is better placed to capture new apps and agents.
Gate questions
- Competitors with >15% share per segment: Security (cloud WAAP) has about 2 to 3 (Akamai, Cloudflare, possibly AWS or Imperva), so it is concentrated but not dominated. Delivery has 3 (Cloudflare, CloudFront, Akamai). In compute, Akamai is not one of the 3 (AWS, Azure, GCP). No segment is fragmented, but Akamai leads none with >30% verified share.
- Replaceable within 12 months? Delivery: yes. Multi-CDN is standard, and the -6% decline reflects this. Security: partially. WAF and bot rules migrate in 6 to 18 months, and Cloudflare is winning share, but large banks and retailers have high switching friction and Guardicore deployments are sticky. Compute: Anthropic is contracted for 7 years, but the commitment is conditional on service levels. Overall: no moat in Delivery, a moderate moat in Security.
- Price setter or taker? Delivery: taker ("Pricing environment remains competitive," Q4 25 call). Security: modest pricing power at the enterprise high end. Compute: it passes through component costs ("we are able to pass on any sort of increase in pricing," Q2 26 call). That is cost-plus behavior, not oligopoly pricing.
Oligopoly hard gate: NOT MET. No segment has a verifiable >30% share. Akamai is one of ~3 leaders in WAAP and CDN, but no source shows those 3 controlling >70% with Akamai clearly at the top. Cloudflare is growing 6x faster. Maximum score is 5.
Durability, buyers, substitutes
- Why durable: Akamai has 25+ years of large-enterprise trust (banks, retail, media, governments) and the most distributed footprint (4,300 locations). It is a WAF incumbent on the largest sites, and outages at competitors reinforce its brand (the Q4 25 call cites a rival's multi-hour holiday outages).
- Buyers: Global 2000 CISOs and infrastructure teams, media and gaming companies, and now AI labs (Anthropic, plus an unnamed tech company with a ~$200M commitment).
- What could replace it: Cloudflare's bundled platform, hyperscaler-native WAF/CDN bundled into cloud spend, and in-house CDNs at the largest traffic generators. Delivery can be substituted today. Security is being substituted slowly.
Assessment
Akamai's thematic exposure is better than its 5% revenue growth suggests. 55% of revenue is in security, growing ~10%, with API security and Guardicore growing faster. CIS is compounding at ~40%, and the verified $11.6B, 7-year Anthropic CPU commitment could make compute 25% to 30% of revenue by 2028. The edge footprint (4,300 locations, 700 cities) is a real distribution asset, even though it is small in GPU or MW terms compared with CoreWeave, Nebius or IREN. Under the investing principles, though, this is not a dominant-leader profile. Akamai co-leads WAAP and CDN with Cloudflare, which is growing much faster. Delivery is a shrinking, price-taking business that a customer can replace within a year. In compute, Akamai rents capacity at sub-1% share and paid a ~5% warrant to win its anchor customer. No segment clears the >30% share test, so the oligopoly gate caps the score at 5. Security leadership and the AI compute ramp keep it from going lower.
Sources: Akamai 8-K Ex. 99.1, Anthropic agreement; TechCrunch, 2026-09-25; Yahoo Finance coverage; Gartner WAAP leaders (SDxCentral); QKS WAAP market share report; wmtips CDN usage; Reanin cloud CDN market; AKAM transcripts FY2025Q4 to FY2026Q2 (FMP, run folder); NET Q2 26 press release (local NET workspace). Data sourced from Daloopa.