Fastly at Citi’s 2026 Global TMT Conference: edge push gains traction

Fastly reported quarterly revenue of $183 million, up 23.3% year over year, with record 65.8% gross margin and roughly 15% operating margin; it has now delivered four consecutive profitable quarters and six quarters of positive free cash flow. Higher-growth security revenue rose 43% and compute/other rose 69%, versus 17% growth in the still-dominant delivery segment, while net revenue retention reached a three- to four-year high of 117%. Management expects security and compute—supported by rising AI and bot traffic—to increasingly drive mix and monetization, although delivery pricing remains down mid- to high-single digits year over year. RPO increased 38% and current RPO 44%, supporting the company’s value-based selling and multi-product adoption strategy.

Analysis

FSLY’s rerating now hinges less on another revenue beat than on whether mix shift can sustain margin expansion while legacy delivery remains structurally deflationary. The market will likely capitalize security/compute revenue at materially higher multiples only if management discloses attach-rate, retention by product cohort, and incremental gross-profit evidence at the September 22 Investor Day. Without those KPIs, the AI narrative risks remaining a low-revenue marketing overlay on a business still dominated by bandwidth-priced delivery.

The underappreciated risk is that the same AI-agent traffic touted as a demand catalyst may be economically unfavorable for CDN revenue: request volumes rise while payload sizes fall, shifting monetization away from the largest revenue line. Cloudflare (NET) is the clearest competitive benchmark; its broader developer/security distribution and larger scale could constrain FSLY’s ability to price security bundles, while Akamai (AKAM) remains a credible enterprise-security alternative. FSLY’s customer concentration makes a single large-account traffic optimization or renewal concession capable of overwhelming several quarters of smaller-logo expansion.

The favorable second-order setup is network utilization: monetizing idle CPU capacity can drive unusually high incremental margins before new infrastructure is required. That advantage is temporary, however; if compute demand becomes material, capex intensity and depreciation will rise, reducing the apparent free-cash-flow conversion investors are extrapolating. After a 181% trailing move, upside requires an Investor Day framework that proves a durable 20%+ growth / expanding-margin profile; absent that, valuation sensitivity to any delivery-growth deceleration is high over the next 1-3 quarters.

Contrarian view: this may be more a successful cost and pricing-discipline turnaround than a scalable AI platform inflection. The right confirmation is not AI traffic growth but security-plus-compute reaching a meaningful share of new bookings, sustained retention above 115%, and delivery price erosion narrowing without sacrificing share.

Key Decisions for Investors

• Maintain a tactical long FSLY only into the September 22 Investor Day if position sizing reflects event risk; take profit on a post-event rally unless management provides segment-level monetization, product attach-rate, and multi-year margin targets. A 3-6 month upside case requires estimate revisions from disclosed booking/retention evidence, not conference commentary.
• Use a pair trade long FSLY / short AKAM for 1-3 months only if FSLY demonstrates security cross-sell acceleration and stable delivery pricing; this isolates the edge-security mix-shift thesis from broad software multiples. Exit if FSLY retention falls below 110%, a top customer drives a material consumption slowdown, or capex guidance rises above the stated range.
• Avoid chasing NET as a direct sympathy trade: FSLY’s claimed AI-security opportunity validates category demand, but NET is more likely to capture it given distribution and product breadth. Prefer NET over FSLY for a 6-18 month structural edge/agentic-traffic exposure if FSLY fails to disclose material paid ContentGuard adoption.
• Set an alert around the next earnings release for delivery growth below the mid-teens, security growth below 35%, or free-cash-flow deterioration despite revenue growth; any of these would challenge the operating-leverage thesis and could trigger sharp multiple compression after the stock’s run.