$GTLB: GitLab Booked 42% Growth and Guided 19%. The Gap Is the Whole Trade.
GitLab beat Q2, raised the year, and gapped 20% to a new 52-week high. The real story is the 23-point spread between bookings growth and revenue guidance, and what closing it in December would be worth.
GitLab ($GTLB) reported second quarter fiscal 2027 results after Tuesday's close, and the market's response was immediate: the stock jumped more than 20% in Wednesday premarket trading to around $54, clearing its prior 52-week high of $52.38. Within the past year, this same stock traded at $18.73.
The headline numbers beat cleanly. Revenue came in at $286.3 million, up 21% year over year, against a Street estimate near $273 million. Non-GAAP earnings of $0.24 per share beat the $0.18 consensus by a third. Management raised full-year fiscal 2027 guidance to $1.129 to $1.133 billion in revenue and $0.85 to $0.87 in non-GAAP EPS, both above where analysts sat.
A beat-and-raise gets you a good day. It does not usually get you 20% and a new 52-week high, especially on a night when enterprise software peers were being punished: MongoDB was down double digits premarket and Credo fell roughly 10% despite its own strong print. Something else moved GitLab. To see it, you need the numbers that were not in the headline.
The quarter under the quarter
For the past two years, GitLab has carried a discount because of a simple, brutal thesis: AI writes code now, GitLab sells seats to people who write code, and fewer people writing code means fewer seats. That thesis is why the stock saw $18 in the past twelve months. Every quarter since, the bears have looked for seat erosion in one metric above all: dollar-based net retention, the measure of whether existing customers spend more or less than they did a year ago.
This quarter, that metric turned. Net retention came in at 117%, the first sequential improvement since 2024. Existing customers are expanding again, not shrinking.
And the expansion was not subtle. Net ARR, the cleanest read on new recurring business booked in the quarter, grew 42% year over year, which management called its second-highest quarterly growth rate in four years, alongside record gross bookings. Deals worth $500,000 or more grew over 150%. First orders roughly doubled to about 1,700 new customers, and new-logo net ARR rose 39%. Customers spending $100K+ annually reached 1,571, up 17%. Public sector demand, soft earlier in the year, rebounded meaningfully.
The AI story flipped from headwind to engine inside the print. Paid consumption of the Duo Agent Platform grew roughly 50% quarter over quarter, and on the call management described one major bank expanding its AI credit pool nearly tenfold. The company also launched GitLab Flex, a consumption-based commercial model, six weeks before quarter-end; over 130 customers had already committed more than $20 million to it by the time the quarter closed.
Briefings like this land in members' inboxes before the market prices them in. Join free
The 23-point problem
Here is where it gets interesting, and where the easy version of this story falls apart.
Bookings grew 42%. Net retention inflected. Large deals grew 150%. And yet the raised full-year guidance implies revenue growth of just 18% to 19%. That is a 23-point gap between the growth rate the sales motion is producing and the growth rate management is telling you to expect in reported revenue.
One of those numbers is what the business is doing. The other is what the income statement will show. The distance between them is an accounting artifact with a specific, quantifiable cause, and it cuts both ways: it flatters nothing today, but it also means the 42% is not fully real cash-in-the-door growth either.
At Monday's close, the market priced GitLab at roughly 5.8 times guided revenue. As of this morning's premarket, that is about 7.3 times. Whether $54 is the start of a rerating or the end of one depends entirely on which growth number you believe, why the gap exists, and what has to show up in the December quarter to close it.
The rest of this briefing is for paid members: the Flex accounting mechanics behind the 23-point gap between bookings and guidance, the real share count and dilution math under the buyback, the December catalyst that settles the argument, and scenario-by-scenario price zones for $GTLB from here.
AlphaBriefing Paid gets you every investment thesis, scenario framework, and catalyst brief we publish, the analysis private intel clients pay four figures for, at a fraction of that.