GitLab (GTLB) is a $955 million revenue DevSecOps platform company that sells a single tool covering the entire software development lifecycle: planning, coding, building, testing, security scanning, deploying, and monitoring. The company is remote-only, operates in 60 countries with roughly 2,580 employees, carries $1.26 billion in cash with zero debt, and just posted its fiscal year 2026 results (ending January 31, 2026) showing 26% revenue growth and 87% gross margins. GitLab went public in October 2021 and trades on the Nasdaq.
The core tension here is straightforward. GitLab has built a genuinely differentiated platform in a market dominated by Microsoft, which owns GitHub, Copilot, Azure DevOps, VS Code, and has a deep partnership with OpenAI. GitLab’s pitch is that enterprises need a single, cloud-agnostic, self-hostable platform rather than stitching together Microsoft’s ecosystem. That pitch has worked well enough to grow revenue from $253 million to $955 million in five years. Whether it continues working as Microsoft bundles AI deeper into GitHub is what determines whether GitLab is worth owning.
How GitLab Makes Money
GitLab uses an open-core model. The free tier gives developers access to the core platform, which includes Git repository management, basic CI/CD, and project planning. This eliminates the adoption barrier and lets teams try GitLab with no upfront cost.
The paid tiers are where the money comes in:
- Premium adds enhanced productivity features, team collaboration, and more sophisticated CI/CD controls
- Ultimate adds everything in Premium plus advanced security scanning, compliance management, supply chain security, and policy enforcement
Both tiers are priced per seat, per year. Customers can deploy GitLab three ways: as a SaaS product (GitLab.com), self-managed on their own infrastructure, or as a fully managed single-tenant instance (GitLab Dedicated, which includes a FedRAMP-authorized option for government customers).
In fiscal year 2026, GitLab introduced two additional monetization layers built around AI:
GitLab Duo is the AI assistant, available in two tiers. Duo Pro provides code completion, generation, chat, and test generation. Duo Enterprise adds security vulnerability analysis, pipeline root cause analysis, and self-hosted model options. Both are seat-based add-ons requiring Premium or Ultimate.
GitLab Credits are the newer and more interesting piece. Credits fund usage of the GitLab Duo Agent Platform, where AI agents autonomously perform tasks across the development pipeline. Unlike seats, credits are consumption-based: customers buy pools of credits that get consumed as agents work. This is GitLab’s first usage-based revenue stream, and it could meaningfully change the growth trajectory if AI agent adoption scales.
Professional services (consulting, implementation, training) round out the revenue mix at roughly 2% of total.
Revenue: Five Years of Compounding
GitLab’s revenue growth has been remarkably consistent, even as the base has scaled:
| Fiscal Year | Total Revenue | YoY Growth |
|---|---|---|
| FY2022 | $252.7M | - |
| FY2023 | $424.3M | 68% |
| FY2024 | $579.9M | 37% |
| FY2025 | $759.2M | 31% |
| FY2026 | $955.2M | 26% |
The five-year CAGR is roughly 30%. Growth is decelerating, which is expected at this scale, but the trajectory has been smooth rather than choppy.
Where the Revenue Comes From
The latest 10-K breaks revenue into four components across three years:
| Component | FY2024 | FY2025 | FY2026 | FY2026 Mix |
|---|---|---|---|---|
| Self-Managed Subscription | $355.7M | $458.9M | $568.5M | 59% |
| SaaS Subscription | $150.6M | $216.3M | $296.2M | 31% |
| Self-Managed License | $63.1M | $68.4M | $68.9M | 7% |
| Professional Services & Other | $10.5M | $15.7M | $21.7M | 2% |
| Total | $579.9M | $759.2M | $955.2M | 100% |
Two things stand out. First, SaaS subscription is the fastest-growing segment at 37% year-over-year, now representing 31% of revenue versus 26% two years ago. Second, self-managed license revenue (recognized at a point in time rather than ratably) has flatlined near $69 million. The business is shifting toward recurring subscription revenue, which now represents 91% of total.
Geographically, the United States accounts for 82% of revenue and that share has been stable across all three years. Europe is 15% and Asia Pacific is 2%. This is heavy US concentration, but it also means international expansion is almost entirely untapped.
Customer Metrics
| Metric | FY2022 | FY2023 | FY2024 | FY2025 | FY2026 |
|---|---|---|---|---|---|
| Customers with >$100K ARR | 492 | 697 | 955 | 1,229 | 1,456 |
| DBNRR | >152% | >130% | 130% | 123% | 118% |
The $100K ARR customer count has nearly tripled in four years, from 492 to 1,456. GitLab also disclosed 155 customers with over $1 million in ARR and a total of 10,682 base customers (those spending more than $5,000 annually).
ARR surpassed $1 billion for the first time during fiscal 2026, with the Ultimate tier representing 56% of total ARR.
DBNRR has declined from over 152% to 118%. This is the metric that bears will focus on, and it deserves attention. A DBNRR of 118% means existing customers grew their spending by 18% on average. That is still healthy expansion, but the downward trend from 152% raises the question of whether this is natural maturation (early customers had more room to grow) or competitive pressure (customers shifting workloads to GitHub). The answer is probably some of both.
Revenue Visibility
As of January 31, 2026, GitLab had $1.1 billion in total RPO (up 20% year-over-year), with current RPO of $719.4 million (up 24%). Combined with $572 million in deferred revenue (cash already collected but not yet recognized), forward visibility is strong.
The Financial Profile
Profitability Trajectory
GitLab’s operating leverage story is one of the most compelling aspects of the business:
| Metric | FY2022 | FY2023 | FY2024 | FY2025 | FY2026 |
|---|---|---|---|---|---|
| Gross Margin | 88% | 88% | 90% | 89% | 87% |
| S&M (% of Rev) | 76% | 73% | 61% | 51% | 46% |
| R&D (% of Rev) | 38% | 37% | 35% | 32% | 29% |
| G&A (% of Rev) | 25% | 28% | 26% | 25% | 20% |
| Operating Margin | -51% | -50% | -32% | -19% | -7% |
Total operating expenses went from 139% of revenue in FY2022 to 95% in FY2026. The company is approaching GAAP operating profitability. Sales and marketing has been the largest source of leverage, dropping from 76% to 46% of revenue as the brand matures and the land-and-expand model takes hold.
Gross margin has drifted slightly lower, from a peak of 90% to 87%. This reflects increasing SaaS hosting costs as more customers move to cloud-hosted GitLab, plus the early impact of AI compute costs for Duo features. This trend is worth monitoring but not alarming at these levels.
Cash Flow
| Metric | FY2022 | FY2023 | FY2024 | FY2025 | FY2026 |
|---|---|---|---|---|---|
| Operating Cash Flow | ($49.8M) | ($77.4M) | $35.0M | ($64.0M) | $232.9M |
| Adjusted FCF | ($53.4M) | ($83.5M) | $33.4M | $120.0M | $219.6M |
| Adj FCF Margin | -21% | -20% | 6% | 16% | 23% |
The FY2025 operating cash flow figure looks anomalous because it includes a one-time $187.7 million tax payment related to a BAPA settlement between the U.S. IRS and Dutch tax authorities. On an adjusted basis, FCF has improved from negative $83 million to positive $220 million in three years.
Stock-Based Compensation
| FY2022 | FY2023 | FY2024 | FY2025 | FY2026 | |
|---|---|---|---|---|---|
| Total SBC | $30.0M | $122.6M | $163.0M | $185.9M | $215.0M |
| SBC % of Revenue | 12% | 29% | 28% | 24% | 23% |
SBC at 23% of revenue is in line with growth-stage software peers. SentinelOne runs at 30%, Figma’s ongoing run-rate is 35-40% once its one-time IPO catch-up charges roll off, and many companies sit in the 20-30% range at this stage of maturity. The trajectory from 29% to 23% over three years while nearly doubling revenue is the right direction. For context, CrowdStrike runs at 23% SBC/revenue at nearly $5 billion in revenue, and Datadog runs at 22% at $3.4 billion. GitLab at 23% on $955 million is not an outlier; it is the industry norm for high-growth security and DevOps software.
The weighted-average diluted share count has grown from 154 million to 167 million over three years, representing roughly 4% annual dilution. In March 2026, the board authorized a $400 million share repurchase program, which should offset two to three years of dilution and signals confidence in the cash generation trajectory.
Balance Sheet
GitLab’s balance sheet is a fortress:
| Item | Amount |
|---|---|
| Cash + Short-term Investments | $1,259.9M |
| Total Debt | $0 |
| Deferred Revenue | $572.1M |
| Stockholders’ Equity | $1,036.2M |
| Goodwill | $17.4M |
Zero debt. $1.26 billion in liquid assets. Only $17 million in goodwill, meaning this growth has been almost entirely organic rather than acquisition-driven. The $572 million in deferred revenue is up 22% from prior year and represents cash already in the bank that will be recognized as revenue over coming quarters.
The company carries $285 million in federal NOLs and $23 million in R&D tax credits, though a $330 million valuation allowance means these are not yet reflected on the balance sheet. As the company turns GAAP profitable, these become a real tax shield.
No single customer represents more than 10% of revenue, though two channel partners account for 30% of accounts receivable, which is a concentration risk worth noting.
The AI Bet
GitLab’s AI strategy is built on three layers, each progressively more ambitious.
The first layer is Duo Pro and Duo Enterprise, which are seat-based AI add-ons. These provide code suggestions, test generation, vulnerability analysis, and chat within the GitLab interface. This is table-stakes AI functionality that competes directly with GitHub Copilot.
The second layer is self-hosted AI models. For enterprises that cannot send proprietary code to external APIs (defense contractors, banks, intelligence agencies), GitLab offers the ability to run AI models on the customer’s own infrastructure through GitLab Dedicated. This is a genuine competitive advantage. GitHub’s Copilot is primarily cloud-based, and many regulated enterprises simply cannot use it.
The third layer is GitLab Credits and the Agent Platform. This is the most speculative but highest-upside component. AI agents that can autonomously plan, write, test, and deploy code would fundamentally change how software teams operate. If a five-person team can accomplish what previously required twenty people, the value GitLab delivers per customer rises dramatically, and usage-based pricing captures that value in a way seat-based pricing cannot.
The risk side of the AI equation is real. GitLab depends on third-party model providers for the underlying AI capabilities. If those providers raise prices, restrict access, or experience outages, GitLab’s AI features suffer directly. AI inference costs are also contributing to gross margin pressure (the 90% to 87% decline partially reflects this), and that pressure will grow as AI usage scales.
Competing Against Microsoft
Microsoft owns GitHub (the world’s largest developer community), Copilot (the first-mover AI coding assistant), Azure DevOps (CI/CD and project management), VS Code (the dominant IDE), Azure (cloud infrastructure), and has a deep partnership with OpenAI. The 10-K names Microsoft explicitly as a competitor and notes that some competitors “may offer their products and services at a lower price or for free.”
This is the single most important risk factor. Microsoft can bundle its entire developer toolchain and undercut GitLab on price. The financial resources mismatch is staggering: Microsoft’s annual R&D budget exceeds GitLab’s entire market capitalization.
GitLab also faces a subtler dependency problem. Many GitLab users work through VS Code, which Microsoft controls. If Microsoft were to restrict or degrade VS Code integrations with non-GitHub platforms, GitLab would lose a key workflow touchpoint.
Given all of that, GitLab has still grown revenue from $253 million to $955 million over five years while competing against this. The reasons it works:
Multi-cloud and vendor independence. Large enterprises running workloads across AWS, Azure, and GCP do not want their developer platform locked to one hyperscaler. GitLab is cloud-agnostic by design.
Self-managed and air-gapped deployments. Defense, intelligence, and regulated financial services organizations need to run development tools on their own infrastructure, often in environments with no internet connectivity. GitHub cannot serve these customers. GitLab can.
FedRAMP and compliance. GitLab Dedicated includes a FedRAMP-authorized offering for federal government agencies. This is a multi-year certification process that creates a real barrier to entry.
Switching costs. Enterprises deeply embedded in GitLab’s CI/CD pipelines, security scanning, and workflow automation face significant migration costs. The 10,682+ customer base and 118% DBNRR suggest these switching costs are real.
Open-core philosophy. Some organizations philosophically prefer open-source tools where they can inspect the code and avoid proprietary lock-in. GitLab’s transparency (including a fully public company handbook) builds trust with this audience.
These factors create a durable niche. GitLab does not need to beat Microsoft broadly. It needs to remain the best option for enterprises that value platform independence, self-hosting, and regulatory compliance.
Valuation
Revenue Projections
Growth has decelerated consistently from 68% to 26%. Management guided FY2027 revenue of $1.099 billion to $1.118 billion, representing 15% to 17% growth. This is a meaningful step-down from FY2026’s 26%, and management acknowledged they are “not satisfied” with this trajectory. The CFO cited two factors: bookings growth that has not kept pace with revenue over the past three years (which flows through mathematically in a ratable model), and roughly 300 basis points of nonrecurring FY2026 tailwinds not embedded in FY2027 guidance. The projections below use management’s guidance midpoint for FY2027 and assume gradual improvement in subsequent years as AI Credits and new go-to-market investments take hold:
| Year | Revenue | YoY Growth |
|---|---|---|
| FY2026 (actual) | $955M | 26% |
| FY2027E | $1,109M | 16% (management guidance midpoint) |
| FY2028E | $1,298M | 17% |
| FY2029E | $1,493M | 15% |
| FY2030E | $1,687M | 13% |
| FY2031E | $1,856M | 10% |
The five-year projected CAGR from FY2026 to FY2031 is roughly 14%. Current RPO of $719 million supports the near-term projection, and the 1,456 customers above $100K ARR provide a large base for continued expansion even at lower DBNRR. Management also guided FY2027 gross margins of 85–87% (down from 89% non-GAAP), reflecting higher costs from SaaS, Dedicated, and the Duo Agent Platform.
Margin Projections
| Metric | FY2026 | FY2028E | FY2031E |
|---|---|---|---|
| Gross Margin | 87% | 86% | 85% |
| GAAP Operating Margin | -7% | +1% | +8% |
| SBC % of Revenue | 23% | 23% | 23% |
| Adj FCF Margin | 23% | 29% | 35% |
Gross margin drifts lower as SaaS hosting and AI compute costs grow. Operating leverage continues as S&M scales from 46% to 32% of revenue. SBC is held flat at 23% of revenue, consistent with where CrowdStrike (23%) and Datadog (22%) run at much larger scale. This means GAAP profitability arrives later and stays modest even at $2 billion+ in revenue.
DCF Valuation
Using SBC-adjusted “true” FCF (adjusted FCF minus SBC) to capture the real economic cost of dilution:
| Year | Revenue | Adj FCF (est.) | Less SBC (23%) | True FCF |
|---|---|---|---|---|
| FY2027E | $1,109M | $277M | ($255M) | $22M |
| FY2028E | $1,298M | $376M | ($299M) | $77M |
| FY2029E | $1,493M | $448M | ($343M) | $105M |
| FY2030E | $1,687M | $540M | ($388M) | $152M |
| FY2031E | $1,856M | $650M | ($427M) | $223M |
Discounting at 10% with a 4% terminal growth rate:
| Component | Value |
|---|---|
| PV of projected FCFs (FY2027-2031) | $413M |
| PV of terminal value | $2,394M |
| Enterprise Value | $2,807M |
| Plus net cash | $1,260M |
| Equity Value | $4,067M |
| Fully diluted shares | ~183M |
| Per share | ~$22 |
Revenue Multiple Cross-Check
Enterprise SaaS companies growing 15-25% with 85%+ gross margins and approaching profitability typically trade in the 8-15x forward revenue range. Applying a range to FY2027E revenue of $1,109 million (management guidance midpoint):
| Multiple | Implied EV | + Cash | Per Share |
|---|---|---|---|
| 8x | $8,872M | $10,132M | $55 |
| 10x | $11,090M | $12,350M | $67 |
| 12x | $13,308M | $14,568M | $80 |
Adjusted FCF Multiple Cross-Check
Using FY2027E adjusted FCF of $277 million (based on guidance-anchored revenue). This is how the market typically values software companies, excluding SBC from the calculation:
| Multiple | Implied EV | + Cash | Per Share |
|---|---|---|---|
| 25x | $6,925M | $8,185M | $45 |
| 30x | $8,310M | $9,570M | $52 |
The gap between the DCF (which penalizes SBC) and these multiples (which exclude it) is the central tension in valuing any high-SBC software company.
Scenario Summary
| Scenario | Revenue Growth Assumption | DCF Value | Multiple Value |
|---|---|---|---|
| Bear | Growth stalls at 10% by FY2029 | $17 | $45 (8x rev) |
| Base | Guidance-anchored, fades to 10% by FY2031 | $22 | $52 (10x rev) |
| Bull | AI re-accelerates growth above 20% | $30 | $67+ (12x rev) |
There is a wide gap between these two approaches. The DCF, which treats SBC as a real cost at a persistent 23% of revenue and anchors near-term projections to management guidance, produces values of $17 to $30. The multiples approach, which is closer to how the market actually prices high-growth software companies, produces $45 to $67+. The adjusted FCF multiples land in between at $45 to $52.
This gap is not a flaw in the analysis. It reflects a genuine disagreement about how to value businesses where 23 cents of every revenue dollar goes to stock compensation. If you believe the market’s convention of ignoring SBC in cash flow multiples is appropriate (and the fact that CrowdStrike, Datadog, and SentinelOne all trade on this basis suggests it is at least standard practice), fair value falls in the $45 to $52 range. If you take the stricter view that SBC is a real and permanent cost to shareholders, the DCF range of $17 to $30 is more appropriate. A balanced view that weights both perspectives puts fair value in the $30 to $45 range.
The largest swing factor remains revenue growth. Management guided FY2027 to just 15–17% growth, a sharp deceleration from 26%. They have acknowledged this is unsatisfactory and outlined five growth initiatives. If AI Credits create a meaningful usage-based revenue stream that re-accelerates growth above 20%, the upper end of the range becomes defensible. If Microsoft’s bundling pressure and SMB weakness push growth into single digits sooner, the lower end is more appropriate. At the current price of $22.26, the stock is trading right at the DCF base case, meaning the market is pricing in very little upside from AI or growth re-acceleration.
What to Watch
| Catalyst | Why It Matters | Timeline |
|---|---|---|
| Q1 FY2027 earnings | First quarter with GitLab Credits revenue contribution; watch for DBNRR stabilization | June 2026 |
| AI Credits adoption metrics | Usage-based revenue is the key growth re-acceleration lever; any disclosure of credits revenue mix is critical | Throughout FY2027 |
| DBNRR trend | If it stabilizes at 115%+ the base case holds; if it drops below 110% the bear case strengthens | Quarterly |
| Share buyback execution | $400M authorized; pace of repurchases signals management’s view of valuation | Quarterly (10-Q filings) |
| Gross margin trajectory | AI compute costs flowing through; needs to hold above 83-84% to sustain FCF margins | Quarterly |
| Microsoft/GitHub competitive moves | Any bundling changes, Copilot pricing shifts, or VS Code integration restrictions | Ongoing |
| FedRAMP and government contract wins | Dedicated deployment for federal customers is a defensible niche; large contract wins would validate the strategy | Ongoing |
Sources:
- 10-K filed March 17, 2026 (FY2026)
- 10-K filed March 26, 2024 (FY2024)
- 8-K filed March 3, 2026 (Q4 FY2026 earnings and FY2027 guidance)
Research and analysis conducted with AI assistance using SEC EDGAR filings as primary sources.