← GitLab Inc.

GTLB · Forward model · Act 2 case

The Act 2 case, 20 quarters out

Model as of

Each segment is projected from its own operating driver, rolled up into consolidated cash flow, and discounted back to a fair value per share. The assumptions below are editable — change them and every number on this page moves with them.

GitLab operates one reportable segment, so there is no segment profitability anywhere in its filings. What it does report, on the face of every income statement since the IPO, is two revenue lines with their own cost of revenue: Subscription - self-managed and SaaS, and License - self-managed and other. Those two lines are this model's verticals, and all twenty-two historical points across eleven quarters are filed figures that sum to reported total revenue to the dollar in every quarter; none is estimated and none is apportioned. Below that line nothing is disclosed: no SaaS-versus-self-managed split, no revenue or ARR for GitLab Duo or the Duo Agent Platform, no geographic revenue, no line-level operating margin, and no total addressable market of any kind. The subscription driver uses Base Customers, GitLab's own defined count of customers paying more than $5,000 of annual recurring revenue, disclosed each quarter for the last seven; the addressable pool of 45,000 is assumed and never binds. The EBITDA margins are calibrated cash margins solved back from reported adjusted free-cash-flow conversion, not reported operating margins, because the engine has no concept of stock-based compensation. The split of GitLab's guided quarter across the two lines applies one implied year-over-year rate to both, which is a modelling choice the company has never made. The $400 million buyback authorised in March 2026 is a financing use of cash with no place in this engine; it appears only in the share count.

Bill Staples' restructuring works on cost before it works on growth. GitLab is cutting 14% of its workforce and exiting 22 countries for $30-35 million of one-time charges, about $19 million of it in the quarter it reports on 1 September 2026. A company running a 14% non-GAAP operating margin that removes a seventh of its people converts materially more of each dollar to cash, even if the disruption costs it some growth for a year - so margin up, growth slightly down, and a multiple between base and bull. What this case does NOT reach is the bull outcome: it produces roughly $1.59 billion of revenue in the twentieth quarter's trailing year against the bull case's $2.13 billion, because cutting cost is not the same as selling agents.

GTLB REVENUE MODEL

Latest: $406M (2032Q1E)

Period Value
2024Q3 $150M
2024Q4 $164M
2025Q1 $169M
2025Q2 $183M
2025Q3 $196M
2025Q4 $211M
2026Q1 $215M
2026Q2 $236M
2026Q3 $244M
2026Q4 $260M
2027Q1 $264M
2027Q2E $273M
2027Q3E $278M
2027Q4E $290M
2028Q1E $290M
2028Q2E $298M
2028Q3E $304M
2028Q4E $317M
2029Q1E $316M
2029Q2E $325M
2029Q3E $331M
2029Q4E $344M
2030Q1E $345M
2030Q2E $354M
2030Q3E $360M
2030Q4E $374M
2031Q1E $374M
2031Q2E $384M
2031Q3E $391M
2031Q4E $405M
2032Q1E $406M

What drives each segment

Subscription - self-managed and SaaS

Subscribers × ARPU
Basis quarter$239M
Final quarter$383M
Implied CAGR+10%
Share of revenue, final quarter94%
PV of segment cash flow$2.41B

90.6% of revenue and effectively the whole equity story. GitLab recognises this line ratably across self-managed subscriptions and its SaaS offering, and it publishes the two things a subscription driver needs: Base Customers, its own defined count of customers paying more than $5,000 of annual recurring revenue, and a dollar-based net retention rate that says what the existing base does to its spending. Both have decelerated for seven straight quarters - customer growth from 16% to 7%, retention from 124% to 117% - while revenue still grew 23%, because the base keeps trading up: $100,000-ARR customers grew 18% and $1 million-ARR customers 26%. So the driver has to carry almost all of its growth in revenue per customer rather than in customer count, and that is exactly what the fiscal 2027 guide implies.

Last four quarters
2026 Q2 $213M Reported
2026 Q3 $223M Reported
2026 Q4 $234M Reported
2027 Q1 $239M Reported
Self-managed subscriptions and SaaS subscriptions to the GitLab DevSecOps platform across the Premium and Ultimate tiers, including GitLab Duo and the Duo Agent Platform, plus professional services. GitLab discloses no revenue, annual recurring revenue or customer count for SaaS versus self-managed, for any tier, or for any AI or agentic product, so nothing inside this line may be sized separately.
Subscribers 11K 24.1% of a 45K addressable base 10,831 Base Customers at 30 April 2026 - GitLab's own defined metric for customers above $5,000 of ARR.
Addressable subscribers 45K the S-curve ceiling Assumed pool of organisations able to pay above $5,000 of ARR. Nothing is disclosed; it never binds inside the horizon.
Net adds 200/qtr ramping toward 120/qtr, throttled as the base approaches the TAM Last four adds ran 234/137/207/149. Against the saturation term this yields about 150 real adds a quarter.
Net-add ceiling 120/qtr what supply can deliver at full rate Adds fade as the base matures and 22 country exits thin coverage. Nothing disclosed sets a floor.
ARPU $7,365.40/mo drifting +1.8% per quarter, floor $0.00 $239.3m of subscription revenue over three months over 10,831 Base Customers. A monetisation index, not a contract value.
Non-subscriber revenue $0/qtr growing +0.0% per quarter Zero: the licence line is its own vertical, so nothing sits outside the Base Customer calculation.
Subscription - self-managed and SaaS

Latest: $383M (2032Q1E)

Period Value
2024Q3 $131M
2024Q4 $142M
2025Q1 $151M
2025Q2 $163M
2025Q3 $175M
2025Q4 $186M
2026Q1 $194M
2026Q2 $213M
2026Q3 $223M
2026Q4 $234M
2027Q1 $239M
2027Q2E $246M
2027Q3E $253M
2027Q4E $259M
2028Q1E $266M
2028Q2E $273M
2028Q3E $280M
2028Q4E $286M
2029Q1E $293M
2029Q2E $300M
2029Q3E $307M
2029Q4E $315M
2030Q1E $322M
2030Q2E $329M
2030Q3E $336M
2030Q4E $344M
2031Q1E $352M
2031Q2E $359M
2031Q3E $367M
2031Q4E $375M
2032Q1E $383M

Assumptions & reasoning

  • Every quarter of this history is a filed number off the face of a condensed consolidated statement of operations, read from seven separate 8-K exhibits in which each quarter appears twice - once as the current quarter and once as the following year's comparative - and every overlap agrees to the dollar. Nothing here is estimated and nothing was apportioned.
  • ARPU here is a monetisation index for the Base Customer cohort, not an average contract value. The line also carries revenue from customers below the $5,000 ARR threshold and from professional services, so this figure is higher than any customer actually pays. It is internally consistent because it is derived from, and reconciles back to, the same disclosed revenue line.
  • The 40.6% EBITDA margin is a CALIBRATED CASH margin, not a reported operating margin, and reading it as one will mislead. GitLab's non-GAAP operating margin was 14% in the basis quarter and 17% for fiscal 2026. Adjusted free cash flow ran far above that - $219.6m on $955.2m of fiscal 2026 revenue, or 23.0% - because roughly $200m a year of stock compensation is non-cash. This engine has no concept for that, so the calibration lives in the margin: solving 23.0% = (E - 1.0% capex) x (1 - 22% tax) gives 30.5% consolidated, adding back the 10.0% corporate overhead line gives 40.5% at vertical level, and splitting that across the two lines by the disclosed gross-margin gap puts this one at 40.6%.
  • The engine's ARPU drift is a single constant. Holding 1.80% a quarter for twenty quarters implies 7.4% a year of expansion forever, which a net retention rate that has fallen one point in each of the last seven quarters would not sustain indefinitely. This is the model's main extrapolation risk, and it is why the bear case moves growth rather than only the multiple.
  • This vertical carries no seasonality. Factors fitted by ratio to a centred four-quarter moving average come out at 0.9877 / 1.0062 / 1.0069 / 0.9992 - a signal of 1.92 points against a worst window-to-window spread of 1.34 points, which is not separable from noise. There is a real mechanism pointing the same way, since GitLab's fiscal Q1 runs 1 February to 30 April and is 89 or 90 days against 92 for the other three, but that alone is worth under a point on a ratable line and is absorbed into the ARPU drift.

License - self-managed and other

Growth path
Basis quarter$25M
Final quarter$23M
Implied CAGR-1%
Share of revenue, final quarter6%
PV of segment cash flow$141M

9.4% of revenue and the lumpy half of the income statement. This line carries the portion of a self-managed term licence recognised up front on delivery, plus other revenue, so it lands in the quarter a contract is signed rather than spreading across it. That makes it visibly seasonal - GitLab's fiscal fourth quarter, its largest renewal quarter, has run about 17% above trend in every year observed - and it makes it the wrong line to drive with customers or ARPU, because GitLab discloses no unit, seat, contract count or price for it anywhere. It grew 7.7% in fiscal 2026 against 26% for the company, then jumped 24% in the basis quarter, and over eleven quarters it has shrunk from 12.5% of revenue to 9.4%.

Last four quarters
2026 Q2 $23M Reported
2026 Q3 $21M Reported
2026 Q4 $26M Reported
2027 Q1 $25M Reported
The up-front recognised portion of self-managed term licences, plus other revenue. GitLab discloses no unit, seat, contract count, renewal cohort or price for this line in any filing, release or exhibit.
Sequential growth -1.2%/qtr decaying toward +1.0% Deseasonalised trend, not a seasonal step. It corrects an unusually strong Q1 and reproduces the guided quarter.
License - self-managed and other

Latest: $23M (2032Q1E)

Period Value
2024Q3 $19M
2024Q4 $22M
2025Q1 $18M
2025Q2 $19M
2025Q3 $21M
2025Q4 $26M
2026Q1 $20M
2026Q2 $23M
2026Q3 $21M
2026Q4 $26M
2027Q1 $25M
2027Q2E $27M
2027Q3E $25M
2027Q4E $31M
2028Q1E $24M
2028Q2E $26M
2028Q3E $24M
2028Q4E $30M
2029Q1E $23M
2029Q2E $25M
2029Q3E $24M
2029Q4E $30M
2030Q1E $23M
2030Q2E $25M
2030Q3E $24M
2030Q4E $30M
2031Q1E $23M
2031Q2E $25M
2031Q3E $24M
2031Q4E $30M
2032Q1E $23M

Assumptions & reasoning

  • Same provenance as the subscription line: eleven consecutive quarters read off the face of the income statement in seven 8-K exhibits, summing with the subscription line to reported total revenue in every one of those quarters, to the dollar. Nothing is estimated.
  • The seasonal factors are fitted by ratio to a centred four-quarter moving average and have a mechanism behind them, not just a shape: up-front licence recognition concentrates in the fiscal fourth quarter, GitLab's largest renewal and booking quarter. The raw series shows it without any smoothing - fiscal Q4 licence revenue of 21.8, 25.9 and 26.1 million against the fiscal Q3 immediately before it of 18.7, 20.8 and 21.1 million in each of the three years observed. The signal is 27.4 points against a worst window-to-window spread of 8.4 points.
  • The weakness worth stating is that only one fiscal Q4 sits inside the centred-average window, so the 1.1734 factor rests on one smoothed observation supported by three raw ones. It is the reason the model needs seasonality at all: without it the January quarter is understated by roughly $5 million and the three around it are overstated.
  • The 25.6% EBITDA margin is the consolidated calibrated 40.5% less the disclosed gross-margin gap between the two lines - 87.2% on subscription against 72.2% here in the basis quarter, a 15.0-point difference reported every quarter on the face of the income statement. Carrying that gap down to EBITDA unchanged is a modelling choice, not a disclosure: operating expenses are shared and GitLab allocates none of them, so no line-level operating margin exists in any filing.
  • GitLab guides only total revenue and has never guided either line. The split of the guided quarter used to calibrate this model applies the same implied year-over-year rate to both prior-year lines, which is an explicit assumption and not something the company has said.
Scenarios

Where each case comes from

Bear case — primary sources

The primary sources this case is built from — filings, calls and posts. Where they are claims by an interested party rather than disclosures, every number in the Bear column is what happens if they are taken at face value.

Bull case — primary sources

The primary sources this case is built from — filings, calls and posts. Where they are claims by an interested party rather than disclosures, every number in the Bull column is what happens if they are taken at face value.

Valuation

From cash flow to fair value

Present value of free cash flow, 20 quarters$1.58B
Terminal-year revenue$1.59B
Terminal-year EBITDA$671M
Exit multiple, on revenue5.0x
Terminal value$7.93B
Discounted at 10.0% a year, terminal value becomes$4.93B
Enterprise value$6.50B
Net cash$1.36B
Equity value$7.86B
Shares0.17B
Fair value per share$46.56
Against the deployed price of $46.54, as of +0%

10% on a business with $1.36 billion of cash and short-term investments, no borrowings of any kind and no financing need. The exit multiple is by far the largest lever here, and it has to be, because 77% of enterprise value sits in the terminal: GitLab trades at 6.5 times trailing revenue today and the base case exits the horizon at 4.5 times, a deliberate 30% derating for a line growing about 9% a year by then and converting roughly 28% of revenue to cash. Holding everything else, 3.5x gives $38.29 a share, 4.0x $41.43, 4.5x $44.57 and 5.0x $47.71 - a full turn of the multiple is worth about $6.30, more than a third of the gap between the bear and the base case. A point on the discount rate is worth about $1.50: 9% gives $46.10 and 11% gives $43.12. The 22% tax rate is GitLab's own stated projected non-GAAP rate for fiscal 2026 and 2027, not an assumption of ours.

Read the other way round: at $46.54 the market is paying 5.0x terminal-year revenue, holding every other assumption on this page fixed. That is the number to argue about.

Quarter by quarter

The projected path

Quarter Subscription - self-managed and SaaSLicense - self-managed and other Revenue YoY EBITDA Capex FCF R40 PV of FCF
2027 Q2E $246M$27M $273M +16% $97M $3M $74M +43 $72M
2027 Q3E $253M$25M $278M +14% $101M $3M $77M +41 $73M
2027 Q4E $259M$31M $290M +12% $106M $3M $81M +39 $75M
2028 Q1E $266M$24M $290M +10% $109M $3M $83M +38 $75M
2028 Q2E $273M$26M $298M +9% $113M $3M $86M +38 $76M
2028 Q3E $280M$24M $304M +9% $117M $3M $89M +39 $77M
2028 Q4E $286M$30M $317M +9% $122M $3M $93M +38 $79M
2029 Q1E $293M$23M $316M +9% $124M $3M $95M +39 $78M
2029 Q2E $300M$25M $325M +9% $129M $3M $98M +39 $79M
2029 Q3E $307M$24M $331M +9% $133M $3M $101M +39 $79M
2029 Q4E $315M$30M $344M +9% $138M $3M $105M +39 $81M
2030 Q1E $322M$23M $345M +9% $140M $3M $107M +40 $80M
2030 Q2E $329M$25M $354M +9% $145M $4M $110M +40 $81M
2030 Q3E $336M$24M $360M +9% $149M $4M $113M +40 $81M
2030 Q4E $344M$30M $374M +9% $154M $4M $118M +40 $82M
2031 Q1E $352M$23M $374M +9% $157M $4M $119M +41 $81M
2031 Q2E $359M$25M $384M +9% $161M $4M $123M +41 $82M
2031 Q3E $367M$24M $391M +9% $165M $4M $126M +41 $82M
2031 Q4E $375M$30M $405M +8% $171M $4M $130M +41 $83M
2032 Q1E $383M$23M $406M +8% $173M $4M $132M +41 $82M

Every row is projected. A year-over-year change is shown only where the quarter it compares against exists — an em dash means there is no comparable quarter, not a flat year.

Track record

Model revisions

Assumptions are marked to reality as each quarter prints. Every change is appended here, with the fair value the model produced at the time, so the model's own history stays visible.

DateFair value thenNote
2026-08-31 $44.57 First model. Built on the fiscal 2027 Q1 basis quarter, one day before GitLab reports fiscal 2027 Q2 on 1 September 2026, so the guided quarter this model reproduces becomes a testable actual almost immediately. The calibration lands Q2 fiscal 2027 at $273.9 million inside the guided $272-274 million, and fiscal 2027 at $1,112.0 million at the bottom of the guided $1,112-1,118 million; the two cannot both be hit at their midpoints by one smooth driver, because the full-year guide implies second-half sequential acceleration that the quarterly guide alone does not.