GitLab shares closed at $45.09 on September 1, up 15-20% in after-hours trading, after the DevOps platform reported Q2 FY27 revenue of $286.3 million — 21% year-over-year growth, roughly five points above company guidance and $13 million ahead of consensus. The print reset a stock that had traded in the high teens since May, when GitLab announced a 14% workforce reduction (about 350 people), an exit from 22 countries, and a 37% geographic footprint cut.

What does the headline beat actually say about the AI-threat thesis?

The 21% growth number is the slowest growth metric in the quarter. Calculated billings grew 24% against 12% the prior quarter. Gross bookings set a company record. And management guided Q3 revenue to $281-283 million, below the just-printed $286.3 million and implying 15-16% year-over-year growth.

The deceleration is intentional. GitLab moved customers onto Flex, which recognizes the license component across the contract term instead of at signing. The CFO quantified the drag at $5 million of revenue pushed out per $50 million of self-managed business converted, with a $13 million maximum full-year impact. Bookings, billings, and cash collection are unchanged. Only GAAP timing shifts.

One caveat the source flags. Both the CEO and CFO told investors "net ARR" grew 42%, and the figure circulated widely. GitLab's investor deck defines eight metrics in its appendix; none is net ARR. With revenue up 21%, net retention at 117%, and $5K+ customer count up 8%, a base growing 42% is arithmetically impossible. The figure reads as net new ARR booked in the quarter versus a year ago — a strong increment, not a base, and one that should never appear next to a 21% revenue growth rate without the label.

How much did Flex actually move in six weeks?

GitLab Flex is a single annual dollar commitment covering Premium and Ultimate seats, GitLab Credits for agent consumption, and new eligible capabilities as they ship. Six weeks in market produced more than 130 customers and more than $20 million in commitments.

The metric management wants investors to track instead of revenue is Paid Consumption Run Rate: Credit commitments, Flex commitments, and paid on-demand consumption, excluding trials and promotional credits. It exited Q1 at $15 million and exited Q2 above $40 million, against a target of more than $100 million by fiscal year end.

The comparable structure is CrowdStrike's Falcon Flex, which crossed $2.29 billion in ARR and 39% of total ARR last quarter on the same commit-and-draw mechanic. Two different companies, one year apart, landed on the same model because the seat stopped being the unit that grows.

Why did gross margin drop 400 basis points in a single quarter?

Non-GAAP gross margin landed at 86%, down from 90% a year ago. Subscription cost of revenue moved from $21.8 million to $38.3 million — 76% growth against 21% revenue growth. Management attributed the compression to SaaS mix shift plus investment in consumption products.

GitLab's nine-quarter margin series — 91%, 91%, 91%, 90%, 90%, 89%, 89%, 88%, 86% — shows five points of compression over two years, with the largest single drop in the quarter the company started selling agent consumption. The source's comparison set is instructive:

• Figma: roughly 500 basis points of gross margin compression, into COGS
• GitLab: 400 basis points, into COGS
• Box: 20 basis points, because Box sells the permissions and audit layer rather than buying inference
• CrowdStrike: subscription gross margin up 100 basis points while capex moved from $30.5 million to $124.4 million, roughly 3% to 8% of revenue

Buying inference lands in COGS and compresses gross margin permanently. Building infrastructure lands in capex and depreciates over years. GitLab is buying — it has integrations with Anthropic's Claude models, Amazon Bedrock, and Google Vertex under an explicit model-neutrality strategy, which makes the inference bill a permanent operating cost.

Is the customer base expanding or just deepening?

The $5K+ customer base reached 11,114, up 8% year over year. The $100K+ base reached 1,571, up 17%. First orders totaled approximately 1,700, more than double a year ago and the largest first-order quarter in three years. Deals of $500,000 or more grew more than 150%.

GitLab added roughly 820 $5K+ customers across the prior full year and landed about 1,700 first orders in a single quarter. The two are not measured the same way — first orders include customers below the $5,000 threshold — but dollars per customer are growing several times faster than customer count, with the top of the base moving up while the bottom stays hard to convert.

Dollar-based net retention printed 117%, which management called the first sequential increase since 2024. The prior two quarters printed 117% and 118%, so on the published figure this reads as stabilization, with any acceleration living in decimals GitLab does not disclose.

Can a 15% non-GAAP margin hold once restructuring rolls off?

GAAP operating margin was negative 20%, against negative 8% a year ago. GAAP operating loss reached $56.9 million. The bridge to non-GAAP is roughly $99 million of adjustments on $286 million of revenue, dominated by stock-based compensation of $75 million (up 38% year over year, now 26% of revenue) and restructuring charges of $19.4 million plus $3.9 million of accelerated stock comp from the May plan.

Operating cash flow was negative $3.1 million, against positive $49.4 million a year ago. Adjusted free cash flow was $9.8 million, a 3% margin versus 20% last year. GitLab still repurchased approximately 3.5 million shares for $104.6 million in Q2 — funded from a balance sheet of roughly $1.3 billion in cash and investments while the stock traded in the twenties and thirties.

Whether the R&D reorganization into roughly 60 autonomous teams produces the productivity gains management is counting on shows up a year from now, not in the next print.