On February 12, 2026, Fastly’s stock did something it has never done before in its seven-year public company history: it surged roughly 40% in a single session. The catalyst was a Q4 2025 earnings report that didn’t just beat Wall Street’s expectations — it demolished them. Revenue of $172.6 million crushed the $161.36 million consensus by more than 7%. The company turned non-GAAP profitable for the first full year. And an AI-driven revenue tailwind that most analysts had barely noticed suddenly became impossible to ignore.
The question now: at ~$2.4 billion market cap and a stock price that’s already reflected a massive re-rating, is the opportunity still there — or did retail investors just buy the peak?
What Fastly Actually Does
Fastly operates a programmable edge cloud platform. In plain English: when you visit a website or use an app, your data doesn’t travel all the way to a central data center and back. Fastly intercepts it at the “edge” — at one of its globally distributed points of presence — delivering content faster, more securely, and with lower latency.
It’s a competitive space. Cloudflare (NET) dominates headlines. Akamai (AKAM) owns the enterprise segment. Fastly carved its niche among developers who need extreme programmability — the ability to run custom logic at the edge, not just cache static files. Its customers include Shopify, Stripe, GitHub, and a growing roster of AI companies.
The platform has three revenue pillars: Network Services (CDN and delivery), Security (Web Application Firewall, DDoS protection, API security), and Compute & Observability (edge computing, Fastly’s longer-term bet). Security and Compute are the high-margin growth engines. Network Services is the cash cow.
The Quarter That Changed the Story
For years, Fastly was the “almost” company. Almost Cloudflare. Almost profitable. Almost worth the valuation. Then Q4 2025 happened.
Revenue: $172.6M — up 23% year-over-year, the fastest growth rate in over three years, and well ahead of the $161.36M consensus. Full-year 2025 revenue came in at $624.0M, up 14.78% from $543.68M in 2024.
Gross margins hit records. GAAP gross margin expanded to 61.4% in Q4 — a new high. Non-GAAP gross margin reached 64.0%. These aren’t incremental improvements — they signal a business whose cost structure is finally maturing.
Remaining Performance Obligations (RPO) jumped to $353.8M, up 55% year-over-year. RPO is essentially Fastly’s backlog — future contracted revenue. A 55% increase means the pipeline is accelerating, not just the near-term numbers.
Free cash flow turned positive: $65.7M for full-year 2025, versus -$10.6M in 2023. The company is no longer burning cash. That’s a material shift for a company that spent years destroying shareholder capital to fund growth.
Non-GAAP net income: $19.7M for full-year 2025 — Fastly’s first full year of non-GAAP profitability since going public in 2019.
One number that remains ugly: GAAP EPS was -$0.83 for the full year, with a GAAP net loss of $121.68M. Fastly still has significant stock-based compensation expense and restructuring charges that keep GAAP in the red. For investors who only look at GAAP, this remains a money-losing business.
The AI Angle: Every Crawler Is a Customer
Here’s what William Blair — the analyst firm that upgraded FSLY to Outperform on February 12 — got right: Fastly is quietly becoming one of the primary infrastructure providers for AI web traffic.
Every time an AI agent (think OpenAI’s web browsing tools, Perplexity, or any of dozens of automated research pipelines) crawls a web page, that request often travels through an edge network. If the target site uses Fastly, Fastly sees that traffic. As AI companies deploy more aggressive and frequent crawlers — and as “agentic AI” systems that autonomously browse the web become mainstream — traffic volumes through edge networks like Fastly’s are structurally increasing.
William Blair analysts noted that “agentic AI traffic is still in its infancy.” The implication: Fastly’s AI-driven volume surge may be early innings, not a one-time bump.
Fastly is leaning into this explicitly. The company launched its Fastly Model Context Protocol (MCP) Server — a product designed to help AI systems interact with Fastly’s platform programmatically. Its Threat Insights Reports on AI crawler traffic suggest the company is building real IP and expertise around this traffic pattern, not just passively benefiting from it.
For Q1 2026, Fastly guided revenue of $168M–$174M with non-GAAP EPS of $0.07–$0.10. The EPS figure is striking: consensus before the print was -$0.03. Fastly guided to profitability in Q1 while the market expected a loss. Full-year 2026 guidance is $700M–$720M in revenue (~14% growth) and $0.23–$0.29 in non-GAAP EPS.
What the Analysts Say — And Why the Divergence Matters
William Blair upgraded from Market Perform to Outperform. Piper Sandler raised its price target from $11 to $14 (Neutral). DA Davidson raised from $9 to $13 (Neutral). The Wall Street consensus remains Hold — six Hold ratings versus two Buys.
The average analyst price target sits at $10.80, which implies roughly 39% downside from the stock’s current price of ~$17.66. That’s an unusual situation: a stock surging on strong fundamentals while the median analyst thinks it’s significantly overvalued.
This divergence tells you something important. The bulls (led by William Blair) believe AI traffic is a structural demand driver that will sustain Fastly’s accelerating growth. The bears argue the stock already reflects a re-rating, and that the underlying competitive dynamics — Cloudflare remains miles ahead — haven’t changed.
The bears have a point. Fastly’s market share in CDN/edge has been roughly flat for years. Cloudflare added features and revenue at a pace Fastly couldn’t match. The 40% single-day pop means a lot of future optimism is now priced in. At ~$17.66 per share with 147 million basic shares outstanding, Fastly trades at roughly 3.8x 2026 revenue guidance — not cheap for a company still running GAAP losses.
What Makes Fastly Different From Cloudflare
The natural comparison is Cloudflare (NET), which is larger (~$40B market cap), faster-growing, and more vertically integrated. But Fastly’s architecture has a key differentiator: programmability at scale.
Fastly was built from the ground up for developers who need to run real code at the edge — Rust, WebAssembly, custom logic — rather than basic caching rules. This makes Fastly the choice for companies with complex edge computing needs: real-time personalization, sophisticated security logic, low-latency APIs.
The problem is that Cloudflare has been building programmability into its platform aggressively, closing the gap. Fastly’s window of differentiation is narrowing, which is why the AI traffic catalyst matters so much — it’s an area where Fastly’s architecture happens to be particularly well-suited.
Key Risks
1. GAAP unprofitability. Fastly burned $121.68M on a GAAP basis in 2025. While non-GAAP profitability is real progress, the gap between GAAP and non-GAAP (primarily stock-based compensation) is large. Share count grew 6.37% in 2025 — dilution is a real cost to existing shareholders.
2. The stock is already re-priced. The 40% single-day surge means most of the easy money from this earnings beat has been made. Buying at $17.66 is buying the reaction, not the surprise.
3. Analyst skepticism. When the median analyst price target is $10.80 against a $17.66 stock, that’s a material divergence. Analysts could raise targets — or the stock could correct. History suggests some reversion is likely.
4. Competitive pressure. Cloudflare continues to outpace Fastly on almost every growth metric. Akamai is entrenched in enterprise. Fastly needs AI traffic to be not just real but large enough to move the needle sustainably.
5. Convertible notes. Fastly raised $180M in 0% convertible notes in 2025 to bolster liquidity and repurchase its 2026 notes. Convertible debt can dilute shareholders at conversion.
The Verdict
Fastly is not the CDN also-ran it was two years ago. The Q4 2025 results are genuinely impressive: record revenue growth, record gross margins, positive free cash flow, non-GAAP profitability, and an RPO backlog that surged 55% — signaling the next several quarters of growth are already contracted.
The AI traffic thesis is real and underappreciated. Every AI agent browsing the web is a Fastly revenue event. William Blair’s framing of FSLY as an “underappreciated AI play” has merit.
But the stock is not cheap after a 40% pop. At ~3.8x 2026 revenue, you’re paying a growth premium that requires the AI tailwind to be sustained — and Fastly to defend against Cloudflare. The median analyst thinks the stock is 39% overvalued right now.
For aggressive growth investors who believe agentic AI will structurally increase edge traffic volumes: FSLY is worth a small position. The RPO growth (+55%) suggests momentum is real.
For value-oriented or risk-averse investors: wait for a pullback toward the $12–$14 range, where the analyst consensus converges. That’s where the risk/reward is more balanced.
Quality of Opportunity Score: 7/10. Great quarter. Credible AI thesis. But the easy trade is already done.
Looking for more small-cap deep dives? Check out our analysis of Mitek Systems (MITK), another AI-adjacent small cap, Magnite (MGNI), a small-cap ad tech play with strong Wall Street backing, and Exodus Movement (EXOD), a crypto-adjacent small cap at an unusual valuation.
Primary source: Fastly Q4 2025 Earnings Release (SEC 8-K via StockTitan). Analyst consensus data via TipRanks.
Disclaimer: This article is for informational purposes only and does not constitute financial advice, investment advice, or a recommendation to buy or sell any security. Margin of Alpha has no positions in FSLY. All financial data is sourced from public filings, earnings releases, and analyst reports. Investing involves risk, including the potential loss of principal. Always conduct your own due diligence before making investment decisions. Past performance is not indicative of future results.