Mike Cikos

Ed described the $1.8B deal as dedicated capacity with ratable revenue recognition once capacity is live. Some revenue in 2026, partial year in 2027 during build-out, then full run rate for remaining years. RPO will show most of it next quarter, building as capacity is delivered. Some nuance in first twelve months relating to goods receipt timing and pricing mechanisms.

Tom said AI advances are making it much better at finding vulnerabilities and enabling attackers. Billions of unpatchable devices are more vulnerable. Application-layer attacks from millions of distributed IPs require a vast distributed platform to defend, WAF in a single data center is insufficient. Akamai's platform with 4,300 locations is needed more than ever. AI helps defense but doesn't solve the problem; net-net a very challenging time for CSOs.

Ed confirmed the $200M deal was an existing customer using Akamai for CDN and security, not one of the largest accounts. Discussions had been ongoing for several months on an exciting workload. Customer dramatically increased spend and Akamai hopes for more business with them.

Ed said capital intensity increase is demand-driven, not structural. The $250M Inference Cloud investment is well-informed by having a large committed customer. No significant change in sourcing complexity, still buying servers and networking equipment. Looking at sourcing from different suppliers to mitigate memory chip cost increases. Still using third-party colo providers. Normalized CapEx excluding memory inflation and Inference Cloud investment would be at low end of typical range.

Tom described it as first inning. Cited commerce companies personalizing with video, AI-powered toys, robotics needing fast edge compute. NVIDIA Blackwell 6000 well suited for edge deployment. Just launching in last couple of weeks with very strong interest.

Ed said formal 2026 guidance will come in February. Only forward-looking comment: tremendous AI Inference Cloud demand creates a very good chance of CIS growth acceleration next year.

Ed confirmed security at about 10% for the year and compute maybe a touch under 15% as larger contracts ramped later than expected. But momentum is clearly accelerating in CIS.

Tom explained CIS is meeting and exceeding expectations with 30% growth on nearly $300M ARR. Full-year compute could land slightly below 15% depending on timing of when large deals start generating revenue. Ed confirmed it comes down to whether customers ramp a month or two late.

Charlie said Evergreen contract length remains 3 to 4 years with no substantial change, and higher prices drive relatively more StaaS demand.

CEO said Evergreen/One orders grew faster than overall revenue and faster than CapEx purchases; the economics are even better in a high-pricing environment (long-term contract, blended cost, buy only what is needed) and he expects it to grow further vs CapEx.

CTO said the first EXA win (a GPU cloud) initially chose an alternate vendor, then ran a performance test, was blown away, and ordered within days with follow-on orders; Everpure is in advanced talks with dozens more.

CEO said Enterprise Data Cloud / Pure Storage Cloud feedback has been extraordinarily positive with rapidly expanding demand; CTO added it is drawing CISOs and CIOs into more strategic conversations.

New CFO said his philosophy doesn't change how Pure guides, it's numbers-driven; the range aligns with industry practice and adds flexibility, with RPO giving comfort on the guide.

CEO said Fusion is part of the array software, so Evergreen//One customers get it almost automatically (Pure manages via Fusion) with high overlap; Portworx is a separate license so overlap varies.

CEO said dozens of customers turned on Fusion within a quarter of launch, a fast uptake with enthusiastic feedback and an expected network effect over several quarters.

CFO confirmed higher-velocity Evergreen//One deals converted to traditional CapEx sales this quarter, lifting product revenue but lowering as-a-service TCV.

Matt Calitri (on for Mike Cikos) asked where customers are finding budget for AI initiatives; Tom sized the enterprise AI market at ~$10B in 2026 growing ~50% and said customers are finding the budgets. Direct.

Matt Calitri (on for Mike Cikos) asked about the public-sector outlook and lingering shutdown impacts; Stephen cited COTS focus, the AI action plan, and maritime reindustrialization as durable federal tailwinds. Direct.

Asked by Matt Calitri on behalf of Mike Cikos. Tom weighted the miss roughly 70% sales disruption and 30% his reduced involvement and compared it to early misses at Oracle and NVIDIA. Direct.

Tom said the deal broadened and extended three years but declined specific economics, citing the NDA. Partial. Matt Calitri asked on behalf of covering analyst Mike Cikos (attribution inferred).

Tom attributed the wider band to real geopolitical, budget, and trade risk and companies withdrawing guidance. Direct. Matt Calitri asked on behalf of covering analyst Mike Cikos (attribution inferred).

Management defended investing for market share and pointed to crossing into cash-positive at some point in fiscal '26. Direct.

Olivier clarified the new lands are neuro labs using Datadog to observe and optimize model training, distinct from the hyperscaler AI labs previously landed, which run similar training workloads but are much larger companies.

David said the share of YoY growth from customers not present a year ago rose from 25% to 30%, reflecting a compounding of both new logo lands and the ramp of logos added over the past year.

Obstler explained they have years of day-adjusted usage analysis experience and incorporate holiday calendar effects, including this year's specific weekday placement, into guidance as standard practice.

Pomel characterized the broad-based non-AI growth as consistent with Datadog's 15-year growth pattern, steady execution across products and go-to-market without a single dramatic catalyst.

Pomel confirmed Flex Logs is now part of all major enterprise deals, resonating strongly in legacy log migration scenarios by changing economics and cost predictability.

Pomel drew distinction: AI natives have exploding traffic-driven growth while enterprises have controlled migration limited by bandwidth. Enterprise bookings are strong but usage growth is more moderate as attention shifts to figuring out AI adoption.

Mike said Atlas should keep rising as a share of revenue but more slowly than expected, since EA is now a more durable contributor with AI demand landing in both.

Mike Gordon said Atlas consumption grew 29% consistent with last year, with notable strength in larger US customers whose workloads grow for longer and expand more than before.

Dev credited the self-serve motion (led by new CMO May Petrie) and its more sophisticated experimentation for the strong new-logo additions.

Mike confirmed May year-over-year consumption returned to February/March levels after an April dip, a direct answer on the recovery.

Serge framed Q4 as the seasonally slowest consumption quarter around the holidays, then pointed to stronger fiscal '25 workload cohorts driving stable FY26 consumption growth.

Michael described the consumption improvement as broad-based while cautioning it was slightly better rather than a step-function change.

Jay said demand very strong in Zero Trust market. SASE term meaningless as every vendor claims it. Platform expansion driving customer base growth. Pipeline and win rate remain strong despite new entrants.

Jay said ZTE customers see 2-3x+ ARR expansion. No per-product quotas but financial incentives for new products and new logos. Workloads expected to grow even as users plateau, creating ongoing upsell.

Jay said macro still tight with large deal scrutiny, but cyber and Zero Trust architecture are high priority. Cost savings messaging helping close deals. Remo praised sales org under Mike Rich for large enterprise focus.

Jay highlighted account-centric selling and GSI growth. 6x upsell opportunity in installed base. GSIs embedding Zscaler into more solutions.

Remo confirmed growth is broad-based across core and emerging. Jay added ZDX, branch, workload, and AI analytics all contributing to emerging products ARR.