Fastly Unveils First Long-Term Targets Since 2023, Collapses Revenue Reporting as Multiproduct Adoption Quadruples
Investor Day, September 22, 2026: Management lays out a path to $1.1-1.3 billion in 2029 revenue while quietly flagging a stubborn new-logo growth problem
Fastly used its first Investor Day in three years to reset the narrative around its edge cloud platform, unveiling 2029 financial targets, a major change to how it will report revenue, and a new AI governance product suite launched just a day earlier. CFO Rich Wong, who joined the company in August 2025, presented a target model calling for revenue of $1.1 billion to $1.3 billion by 2029, implying a 14% to 21% compound annual growth rate from 2026, alongside gross margins expanding to 67-71%, operating margins reaching 20-22%, and free cash flow yield of 12-15%. The targets mark the company's first refreshed long-term guidance since its 2023 Investor Day and come after three consecutive quarters of 20%-plus year-over-year growth and six straight quarters of positive free cash flow.
Revenue Reporting Overhaul Signals Platform Bet
The most structurally significant announcement was Fastly's decision to abandon its three-line revenue reporting structure of network services, security, and other, moving to a single consolidated revenue line beginning in 2027, supplemented by a new multiproduct adoption metric tracking the percentage of customers buying two or more, and four or more, of the company's six product suites. Wong will run both reporting methods in parallel for the next two quarters to ease the transition. His rationale was blunt: "We have one network... and being one network, we're running kind of like delivery and security, and we're kind of providing intelligence on that delivery traffic, whether that intelligence is security or compute, but it becomes a very blurred line. And I think that blurred line becomes really messy and kind of misleading sometimes." CFO Wong argued that new products like AI Runtime Control don't fit cleanly into legacy categories, making the old lines increasingly arbitrary as the platform strategy deepens. Analysts on the call, including Citi's Fatima Boolani, pressed management on why the security line was being retired just as it becomes central to the growth story; the company's answer effectively signaled that going forward, investors will need to track platform-wide multiproduct adoption rather than segment-level detail.
Multiproduct Adoption Is the Central Financial Story
The number management wants investors to fixate on is multiproduct adoption: 72% of customers now use two or more of Fastly's six product suites, up from 58% two years ago, while 30% use four or more suites, a figure that has roughly doubled each year, up from just 7% two years ago. CEO Kip Compton and Chief Product Officer Kelly Shortridge both framed this as evidence that the company's historically fragmented, single-product sales motion has been replaced by a genuine platform sale. Shortridge, who has led Fastly's product organization since November 2025 after previously running the security business to 53% growth, noted that over the last four quarters, half of all new business deals included a security product, calling security "that new on-ramp" into the broader platform. Security revenue itself grew 43% year-over-year last quarter, with DDoS protection and Bot Management both growing in the triple digits.
New Logo Growth Has Stalled, Even as Top Accounts Surge
Beneath the platform narrative sits a less comfortable data point. Analyst Vijay Homan of Craig-Hallum noted that net new customer growth ran around 17% annually in Fastly's first three years post-2023, but has slowed to roughly 2% over the last two years. President of Go-to-Market Scott Lovett attributed this partly to the historical friction of enterprise migrations, arguing that AI-assisted configuration tools have now cut migration timelines from weeks to as little as two days, which he expects to unlock faster new-logo acquisition going forward. Meanwhile, the top 10 customers grew 48% year-over-year last quarter, compared to just 12% for the non-top-10 base, a gap Wong acknowledged is "not where we want it to be." The concentration risk is real: Fastly's growth story over the past two years has been driven overwhelmingly by deepening wallet share with its largest, most sophisticated accounts, eight of the ten largest streaming platforms and six of the seven largest technology companies among them, rather than broadening the customer base.
Tightened Incremental Margin Model
Wong also refined the incremental margin framework Fastly introduced in 2023. The company narrowed its incremental gross profit flow-through range from 65-80% to 70-80%, and its incremental operating income flow-through from 25-40% to 30-40%. On a last-12-month basis, Fastly has already been running at 96% incremental gross profit flow-through and 79% operating income flow-through, well above the new targeted ranges, giving Wong room to argue the tightened model is conservative rather than aspirational. He attributed the improvement to better network utilization forecasting, tighter deal and pricing discipline, and a richer mix of higher-margin products layered onto the existing network.
Network Efficiency Is the Underappreciated Moat
Founder and CTO Artur Bergman and Compton both emphasized that Fastly's single, hyper-converged network, the same infrastructure simultaneously serving Thursday Night Football streaming, Black Friday e-commerce traffic, and video game launches, is driving structural cost advantages over competitors who run separate infrastructure stacks for delivery, security, and compute. The company disclosed it grew network capacity 65% while cumulative capital expenditure over that period totaled only around 10% of revenue, a ratio Wong called out as a key reason gross margins expanded from 59% to 66% year-over-year even as GPU and hardware costs rose industry-wide. Notably, Fastly's deliberately small footprint of 166 points of presence, far fewer than most competitors, was framed not as a weakness but as evidence of architectural discipline; Compton recounted a customer story in which a major retailer's Icelandic traffic, moved to Fastly despite the company having no local point of presence there, outperformed a competitor that did have local infrastructure.
New AI Governance Products and the "Cascade Inference" Bet
A day before the Investor Day, Fastly announced three new AI-focused products, AI Runtime Control, an AI Firewall, and API Enforcement, its first formal entry into what management calls "governing AI." Shortridge described the strategic logic as pairing Fastly's existing compute product with the new runtime control layer to create what the industry calls cascade inference, where small, efficient models handle the vast majority of requests at the edge and only escalate to large, expensive frontier models when necessary. "That cascading inference, that symbiosis between compute and AI runtime control is what makes AI economically viable to adopt at scale for enterprises," Shortridge said, positioning this as merely "step one" of a much larger platform-wide AI opportunity that touches security, delivery, and compute simultaneously. Bergman was explicit that Fastly has no ambition to run frontier-scale large language models at the edge, arguing GPUs remain poorly suited to multi-tenant edge economics today; instead the company is betting on small, cheap models, citing his own $100 home project training a sub-4-megabyte model to detect children near a pool, and a colleague's 3.3-megabyte chess model that reportedly beats him consistently, as evidence of what's technically and economically viable at the edge right now.
Early but Accelerating AI Traffic Signals
Management shared granular data points on AI-driven traffic patterns without providing a specific revenue attribution. Fastly said it has seen 11-times growth in Model Context Protocol traffic year-to-date, a 34-times increase in code-generation traffic in just the last couple of months, a 12-times increase in content-authenticity traffic, and a 7-times increase in AI model distribution across the platform in the past month alone. Bergman highlighted Python Package Index download data showing a sharp inflection coinciding with the release of Anthropic's Claude Code, illustrating what he called agentic rather than human-driven traffic growth: "That is not human behavior that is driving the increase. That is agentic behavior." He described this as a "tsunami of traffic" that has effectively "disconnected infinite usage from human time," since AI agents operate continuously rather than within human waking hours. Compton was careful to note the company cannot yet isolate a discrete AI revenue bucket, but argued rising traffic volume, new value-added products like bot and DDoS protection benefiting from agentic traffic growth, and eventual edge inference monetization together represent three distinct paths to converting the traffic surge into revenue.
International Expansion, Led by a Striking India Data Point
Lovett disclosed that Fastly's points of presence in India expanded from 12 to 62 in roughly seven months through a partnership embedding its software onto a local service provider's infrastructure, a scale-up he said stopped enterprise customers mid-presentation during a recent trip with Bergman. The company is also adding sales resources and points of presence in Latin America this year, having hired a new regional vice president, and plans further expansion into the Middle East, Thailand, and Southeast Asia in 2026. Lovett was candid that the era of traditional reseller partnerships is over, telling the room that "the days of resale... it's really a services-led play right now," with partners increasingly expected to provide integration expertise rather than simply fulfill demand.