MongoDB [MDB]
As of 2026-07-29. Spot $320.98. Framework v1.5.1. Tier-2 memo, cluster member (data & dev infrastructure).
Cluster analysis: reports/clusters/Data_Dev_Infrastructure_Cluster_2026-07-29.md.
This memo issues no position verdict. Every Criteria returns PASS / FAIL / INDETERMINATE. The book decides.
The brief warned that net cash was wrong on 11 of 11 names last night. It is not wrong on MongoDB, and that is worth recording as precisely as the errors are.
| Input | Screen | Verified (10-Q for quarter ended 2026-04-30, filed 2026-05-29) | Status |
|---|---|---|---|
| Shares outstanding | 80,431,927 | 80,503,576 (balance sheet @2026-04-30); dei 80,431,927 @2026-05-27 | CORRECT |
| TTM revenue | $2,602,399k | $2,602,399k (FY26 $2,463,797 − Q1 FY26 $549,014 + Q1 FY27 $687,616) | CORRECT |
| Latest-quarter YoY | +25.2% | +25.25% ($687,616 / $549,014) | CORRECT |
| Net cash | $2,377,878k | $2,371,924k like-for-like — a −$5.95m, 0.25% difference | CORRECT |
| GAAP operating margin | −5.6% | −4.16% TTM (screen used FY2026, one quarter stale) | STALE, 1.4pp |
| Gross margin | 71.7% | 72.2% latest quarter | STALE (understated) |
| EV / Sales | 8.68x | 9.00x (on today's higher spot) | consistent |
| Terminal margin | 10.7% | Not name-specific — see §7 | DEFECT |
| Component | @2026-04-30 | In screen? |
|---|---|---|
| Cash and equivalents | $1,036,354k | yes |
Short-term investments (AvailableForSaleSecuritiesDebtSecuritiesCurrent) |
$1,390,799k | yes |
| Restricted cash, non-current | $3,406k | no (immaterial) |
| Finance lease liabilities (current + non-current) | $(28,208)k | yes |
| Operating lease liability, non-current | $(21,067)k | yes |
| Operating lease liability, current | $(9,360)k | no (immaterial) |
| Convertible debt | $0 | — |
| Net cash, like-for-like | $2,371,924k | vs screen $2,377,878k |
MongoDB has no convertible notes and no debt. The $1,142–1,145m of 0.25% Convertible Senior Notes due 2026
were retired: ConvertibleDebtNoncurrent went to zero at 2024-10-31 and has stayed there. There were no
non-current marketable securities to omit. The screen's arithmetic was correct because there was nothing left
out. That is the material point — the net-cash routine is not universally broken; it fails when a name has
non-current investments (SNOW), off-pattern tag names (GTLB) or convertibles (SNOW, DDOG). MongoDB has none of
those.
Preferred net cash for valuation: $2,430,559k (cash + short-term investments + restricted cash; leases excluded as operating obligations, applied consistently across the cluster). EV = $25,840m − $2,431m = $23,410m. EV/Sales = 9.00x.
Q1 FY27 net income $4,434k ÷ 80,357,498 basic weighted-average shares = $0.0552, against a filed $0.06 basic and $0.05 diluted ($4,434k / 81,581,387 diluted = $0.0543). Scale confirmed on both lines.
This is also MongoDB's first quarter of positive GAAP net income in its history as a public company — the annual series runs −$345.4m (FY23), −$176.6m (FY24), −$129.1m (FY25), −$71.2m (FY26), and now +$4.4m for a single quarter.
Filing recency: 10-Q filed 2026-05-29. 61 days old. Fresh.
Verdict: yes. This is the cleanest revenue line in the cluster.
Quarterly revenue and YoY:
| Quarter ended | Revenue | YoY |
|---|---|---|
| 2025-04-30 | $549.0m | +21.8% |
| 2025-07-31 | $591.4m | +23.7% |
| 2025-10-31 | $628.3m | +18.7% |
| 2026-01-31 | $695.1m | +26.7% |
| 2026-04-30 | $687.6m | +25.25% |
Stable in the low-to-mid twenties, with the January quarter as the seasonal peak. No acceleration to explain and no deceleration to worry about — the screen's STABLE flag is correct.
MongoDB attributes the growth specifically: "Subscription revenue increased by $134.7 million primarily due to an increase in consumption of Atlas by our large existing customers as evidenced by our net ARR expansion rate of 121%."
Accounting quality checks: - One acquisition, in the FY2026 year, and it is small. Goodwill rose from $69,679k to $191,397k during FY26, of which $121,718k was "increase in goodwill related to business combinations." No goodwill was added in Q1 FY27 ($191,397k at both 2026-01-31 and 2026-04-30), so the +25.25% latest quarter contains no first-time acquisition contribution. Total intangibles are $51.8m gross / $30.9m net — de minimis on $2.6bn of revenue. - No settlement, milestone or one-off revenue. - No customer concentration disclosed above the 10% threshold. - Receivables are behaving. $499.0m (Jan-26) → $387.3m (Apr-26), a $111.7m collection. DSO on the latest quarter's revenue is 51 days — the lowest in the cluster and down from 65 days at January. No DSO stretch of the Applied Optoelectronics class. - Accruals clean. TTM operating cash flow $596.9m against a TTM net loss of $(66.7)m — a $664m gap fully explained by SBC of $555.9m plus D&A and working capital. Cash generation is real and growing fast: FY24 $121.5m, FY25 $150.2m, FY26 $505.1m, Q1 FY27 $201.6m (against $109.9m in Q1 FY26, +83%).
| Quarter ended | Revenue | Gross profit | Gross margin |
|---|---|---|---|
| 2025-04-30 | $549.0m | $391.0m | 71.2% |
| 2025-07-31 | $591.4m | $420.0m | 71.0% |
| 2025-10-31 | $628.3m | $449.1m | 71.5% |
| 2026-04-30 | $687.6m | $496.2m | 72.2% |
Gross margin is up 100bp YoY, not down. Unlike GitLab, MongoDB's mix shift toward cloud-hosted Atlas has not compressed margins over the last year. MongoDB does flag the structural risk ("cost of subscription revenue may increase as a percentage of subscription revenue" as Atlas grows), and at 72.2% MongoDB has the second-lowest gross margin in the cluster — a database-as-a-service carries real infrastructure cost. But the direction over the last four quarters is favourable and the level is stable.
MongoDB is the purest consumption business here, and its own disclosure says so.
| Value | |
|---|---|
| Remaining performance obligations @2026-04-30 | $1,458.6m |
| RPO one year earlier | $774.3m — +88.4% YoY, the fastest in the cluster |
| Expected recognition in next 12 months | 53% (disclosed) |
| Expected in 13–36 months | 46% (disclosed) |
| Committed forward-12-month revenue | $773m |
| As % of TTM revenue ($2,602.4m) | 29.7% — the lowest in the cluster |
| Deferred revenue (current + non-current) | $432.3m |
(a) The practical expedient. MongoDB states: "The Company applies the practical expedient to omit disclosure with respect to the amount of the transaction price allocated to remaining performance obligations if the related contract has a total duration of 12 months or less." So the $1,458.6m systematically excludes MongoDB's short-duration business — which is a large share of it. The 29.7% is therefore a floor, not the true committed share.
(b) The consumption disclaimer, verbatim: "the amount and timing of revenue recognition are generally dependent upon customers' future consumption, which is inherently variable at the customers' discretion." This is the same language SNOW carries.
And MongoDB is explicit that the direction of travel is toward less commitment:
"Customers sold by our sales force may also sign contracts that remain in effect until terminated and are invoiced monthly in arrears based on usage. We expect to continue to see a higher portion of our Atlas contracts to be billed monthly in arrears based on usage without requiring upfront commitments."
MongoDB is telling readers its committed share will fall. That is an important, forward-looking statement and it is the opposite of what a five-year underwriter wants to hear.
The +88.4% RPO growth deserves attention. The step happened in one quarter: $994.3m (Oct-25) → $1,472.7m (Jan-26), a $478m jump, then $1,458.6m in April. That is a very large multi-year commitment or small number of commitments signed in the January quarter. MongoDB does not identify it. It is a genuine positive — commitments are growing far faster than revenue — but it also means the RPO base is now less diversified than the revenue base, and the 46%-in-13-to-36-months bucket is concentrated in whatever was signed then.
Is a five-year implied path underwritable? Marginally. Only 29.7% is contracted at the 12-month horizon, and management expects that share to fall. But the uncommitted majority is usage of a production database, which is the stickiest form of consumption in this cluster: a workload does not stop consuming without an application being decommissioned. That is a qualitative argument, not a contractual one, and it should be labelled as such. Supporting evidence for the stickiness: customers with $100,000 or more of ARR grew from 2,506 to 2,895 (+15.5%) — an independently-counted, growing base of large accounts.
Form: an exact integer percentage, current period only. No prior-year comparative is presented in the 10-Q.
"As of April 30, 2026, our net ARR expansion rate was 121%."
| As of | 2026-04-30 | 2025-04-30 |
|---|---|---|
| Net ARR expansion rate | 121% | not disclosed in this filing |
MongoDB's definition is given in full and is the stricter of the two forms in common use: "dividing the ARR at the close of a given period ... from customers who were also customers at the close of the same period in the prior year ... by the ARR from all customers at the close of the base period, including those who churned or reduced their subscriptions." That denominator includes churn, which makes 121% a harder number than a same-cohort-only calculation would produce.
Would deterioration be detectable? Partially. The value is exact, so a fall from 121% to 114% would be visible next quarter to a reader who recorded 121% this quarter. But MongoDB presents no comparative and no multi-period table, so the reader must construct the series himself from successive filings. Contrast SNOW, which publishes a five-quarter table (124/125/125/125/126), and GTLB, which puts the prior year in the same sentence (117% and 122%).
This is the middle of the cluster's five disclosure forms, and the risk it creates is specific: a slow drift — 121% → 119% → 117% → 115% — is fully visible to a diligent reader and completely invisible to anyone reading a single filing. It is not the TTAN censoring problem, but it is not SNOW's transparency either.
No gross retention rate is disclosed, so churn cannot be separated from expansion within the 121%. MongoDB does disclose the tracked-metric list ("ARR, net ARR expansion rate, total customers, direct sales customers, Atlas customers, customers over 100K and downloads") and a standard caveat that operational metrics are "subject to inherent challenges in measurement."
Supporting metric, disclosed and exact: customers with $100,000 or greater in ARR: 2,506 → 2,895, +15.5%. Against revenue growth of 25.25%, that gap is expansion within the large-customer base, which is consistent with a 121% expansion rate.
| Value | |
|---|---|
| SBC, TTM | $555,853k |
| as % of TTM revenue | 21.4% |
| GAAP operating loss, TTM | $(108,218)k |
| SBC ÷ |GAAP operating loss| | 5.1x |
| Buybacks, TTM | $500,588k ($400,333k in FY26 + $100,255k in Q1 FY27) |
| Buyback ÷ SBC | 90% |
| Shares outstanding, YoY (dei cover page) | 81,714,313 → 80,431,927 = −1.6% |
| Basic weighted-average shares, YoY | 81,061k → 80,357k = −0.9% |
Is GAAP operating margin positive? No: −4.16% TTM. But the improvement is steady and large: FY2023 −27.0%, FY2024 −13.9%, FY2025 −10.8%, FY2026 −5.6%, TTM −4.16%. That is +23pp over three years and +1.4pp in the last quarter alone.
The share count fell. In this cluster only MongoDB and Atlassian can say that, and MongoDB's decline is the larger on a cover-page basis: −1.6% YoY. MongoDB began repurchasing in FY2026 — $400.3m in the year, $100.3m in Q1 FY27 — and at $500.6m against $555.9m of SBC it recovers 90% of the dilution. Against SNOW's 42% and Datadog's 0%, that is a materially different capital-allocation posture.
Is the terminal margin assumption near reality? Closer than for SNOW or DDOG, and short of GTLB and TEAM. The screen assumed 10.7%. MongoDB's non-GAAP operating margin — GAAP −4.16% plus SBC 21.4% — is approximately +17.2%. §7 shows the price requires a 23.6% terminal EBIT margin at a 20x exit — 6.4pp above the current non-GAAP margin. That is a real gap but it is a plausible gap: it requires roughly 6pp of operating leverage on a business already expanding margin 5–7pp a year, with no requirement that SBC collapse. Compare SNOW (needs 36.2% from 6.1%) and DDOG (needs 49.1% from 20.6%).
Where the free cash flow now is: TTM operating cash flow $596.9m on $2,602.4m of revenue = 22.9%, with capital expenditure of $7.3m TTM. MongoDB converts 22.7% of revenue to free cash flow while reporting a GAAP operating loss — the entire gap is SBC, and 90% of the SBC is being bought back with the cash. That is the cleanest cash-versus-GAAP reconciliation in the cluster.
The named mechanism: consumption growth of Atlas within large existing customers, measured by a 121% net ARR expansion rate and a 2,895-strong base of $100k+ ARR customers, plus continued displacement of MongoDB Enterprise Advanced (self-managed term licences) toward Atlas (database-as-a-service).
The mechanism is evidenced by two independently-counted disclosures rather than by narrative: the expansion rate and the $100k+ customer count (+15.5%).
AI monetisation: narrative, and thinner than any other name in the cluster. MongoDB names vector search and AI application workloads. No AI revenue, no AI ARR, no vector-search adoption figure. Not credited.
| Term | FY2025 10-K | FY2026 10-K | Change | Apr-2025 10-Q | Apr-2026 10-Q |
|---|---|---|---|---|---|
| "Atlas" | 102 | 96 | −6% | 62 | 67 |
| "vector" | 6 | 10 | +67% | 1 | 2 |
| "agent" | 5 | 10 | +100% | 5 | 7 |
| "consumption" | 3 | 5 | +67% | 2 | 4 |
| "net ARR expansion" | 8 | 8 | flat | 6 | 7 |
| "Enterprise Advanced" | 16 | 14 | −13% | 8 | 8 |
| "optimiz-" | 3 | 3 | flat | 1 | 1 |
(Methodology: term frequency in the primary 10-Q/10-K documents; transcript counts were not obtained in this run's time box. Like-for-like comparisons only. The bare token "AI" is excluded — it substring-matches inside ordinary words and produced meaningless counts of 700–1,200 per filing.)
Three readings.
Product-cycle facts, dated and from the filings: MongoDB Enterprise Advanced remains the self-managed commercial product and its mention count is falling (16 → 14), consistent with the Atlas mix shift. Atlas was introduced in June 2016 and the company states self-serve Atlas customers "are charged monthly in arrears based on their usage" — the mix shift is toward less committed revenue (see §3). Intangibles acquired comprise developed technology ($27.4m gross) and IP addresses ($24.4m gross) — an unusual asset class, and the $121.7m of FY26 goodwill is not attributed to a named target in the quarterly filing.
| Criteria | Type | Result | Basis |
|---|---|---|---|
| Quality | BINDING | PASS | Archetype INFLECTION. Gross margin (LEVEL) 72.2% — PASS, and improving 100bp YoY, the only name in the cluster where gross margin is rising. Operating margin (CHANGE): −27.0% (FY23) → −13.9% (FY24) → −10.8% (FY25) → −5.6% (FY26) → −4.16% (TTM). +23pp over three years, +6.6pp over the last two — above the ~+5pp separation. Revenue growth: level 25.25%, above the ~18% standard; acceleration approximately flat (21.8% → 25.25% over five quarters, +3.5pp) — passes on both legs of the disjunction. F-score and gross-profitability levels demoted to context per the INFLECTION standard. Accruals: the cleanest in the cluster — TTM operating cash flow $596.9m against a net loss of $(66.7)m, DSO down from 65 to 51 days, receivables down $111.7m in the quarter. No accrual quality concern of any kind. |
| Valuation | BINDING | PASS | See §8. Requires 12.6% CAGR at the 8.0x cluster base against 25.25% demonstrated: +12.6pp margin. Passes at 4x (29.4% required vs 25.25% — marginal FAIL), 6x, 8x, 10x and 12x. Flagged: at a 4.0x exit it requires 29.4% and would FAIL, so unlike GTLB and TEAM the result is not robust across the whole range — but 4.0x is 55% below MongoDB's current 9.00x and 67% below its post-2023 median, so it is an implausible corner. The verdict holds from 6x upward. |
| Downside | MEASURED | scored, see §9 | Named cause: workload consolidation onto hyperscaler-native databases. |
| Liquidity | BINDING | PASS on equity, WEAKEST OPTIONS IN THE CLUSTER | $25.8bn market cap — equity liquidity is not in question. Options chain pulled, Jan-2027 calls, and they are the poorest in the cluster: $320 strike OI 477, bid $63.35 x46 / ask $71.36 x26 — a 12.6% bid-ask spread with only 46 bid and 26 offered; $340 OI 324, x34 / x41; $360 OI 367, bid x8 / ask x11 — eleven contracts offered; $400 OI 359, x38 / x16. IV 73–75%. MDB carries the highest implied volatility in the cluster after TEAM and the thinnest quoted size. This is not the HCA failure (18 contracts of maximum OI across an entire chain) but a defined-risk spread at $340/$400 would face 12% of slippage on entry and quoted size in the tens of contracts. Any options structure on this name must be sized to the quoted depth, not to the open interest. The equity is the appropriate vehicle. |
| Momentum | MEASURED | positive, second-best in the cluster | 12-1 momentum: $238 (Jul-25) → $336 (Jun-26) = +41.2%, second only to DDOG. Path: $210 (Jun-25) → $420 (Dec-25) → $245 (Mar-26) → $336 (May-26) → $321 (Jul-26). More two-sided than DDOG's — MDB has made round trips in both directions rather than one vertical month, which makes its 12-1 signal better-founded. Cross-sectionally a positive-momentum name. |
| Catalyst | MEASURED | scored | Q2 FY27 results, expected late August 2026 (pattern: 2025-08-26, 2024-08-29, 2023-08-31). See catalyst calendar. |
| Consensus | MEASURED | INDETERMINATE | Alpha Vantage quota exhausted. Blocks nothing. |
| Short Mechanism | MEASURED | absent | The test is decelerating growth and exhausted margin runway. Neither leg is present: growth is stable-to-modestly-accelerating (21.8% → 25.25%), and margin runway is large and being spent productively (+23pp in three years, 21.4% of revenue still in SBC). No short mechanism. |
| Peer Spread | MEASURED | scored | Named peer in the same end-market: SNOW (consumption-priced data platform) and, as the competitive threat, the hyperscalers' native document databases (Amazon DocumentDB, Azure Cosmos DB). MDB 9.00x EV/S at 25.25% growth vs SNOW 19.07x at 33.5% — MDB trades at a 53% discount to SNOW for 75% of the growth rate. Own-history: 21st percentile post-2023-01, 36th post-2024-07. |
| Sub-sector | MEASURED | Data & Dev Infrastructure — consumption |
Full working in MDB_Valuation.md.
Implied-path test (Valuation Criteria). Solved on reverse_dcf.py, EV/Sales basis, held fixed: 5-year
horizon, WACC 10.0%, exit multiple as shown, net cash $2,430.6m, shares 80.504m, TTM revenue $2,602.4m.
| Exit EV/S | 4.0x | 6.0x | 8.0x (base) | 10.0x | 12.0x |
|---|---|---|---|---|---|
| Required revenue CAGR | 29.4% | 19.3% | 12.6% | 7.7% | 3.9% |
| Margin vs 25.25% demonstrated | −4.2pp | +6.0pp | +12.6pp | +17.6pp | +21.4pp |
Implied multiple compression at the base: 9.00x → 8.0x = −1.00x, an 11% compression. The smallest compression of any of the three richly-valued names in this cluster, and the reason MongoDB's margin is comfortable while SNOW's is not: MongoDB is only 11% above the base exit multiple, while SNOW is 138% above it and DDOG 208% above it.
Required terminal EBIT margin at each exit multiple, growth held at 25.25%: 31.4% at 15x / 23.6% at 20x / 18.9% at 25x / 15.7% at 30x. Current non-GAAP operating margin ≈ +17.2%. At a 25x exit the requirement (18.9%) is within 1.7pp of what MongoDB earns today before SBC.
12-month target: $437, +36.1% to spot — the largest upside in the cluster. - NTM revenue $3,227m, a house estimate at 24% growth — a touch below the 25.25% latest quarter, consistent with a stable trend and the law of large numbers. - Multiple 10.9x P/S = the post-2024-07 median of MongoDB's own history. Current 9.93x sits at the 36th percentile of that window and the 21st of the post-2023 window. The anchor is IDENTIFIED, unlike GTLB's and TEAM's — MongoDB's post-2024 series is genuinely two-sided (10.4 → 15.3 → 6.3 → 15.3 → 9.9), with multiple round trips in both directions, so median reversion is a legitimate 12-month anchor rather than the midpoint of a de-rating. - The 2021 bubble and 2022 crash are excluded as a regime change; the full-history median of 13.0x is reported for completeness but not used.
No external professional target available for the sanity band.
Type MEASURED. Logged, scored, does not reject the name.
Named cause: workload consolidation onto hyperscaler-native document databases, on a revenue base that is 29.7% contracted and moving toward monthly-in-arrears billing with no upfront commitment.
The mechanism is structural and MongoDB supplies both halves of it. First, only 29.7% of forward-12-month revenue is contracted, and management states it "expect[s] to continue to see a higher portion of our Atlas contracts to be billed monthly in arrears based on usage without requiring upfront commitments." The committed share is falling by design. Second, MongoDB's own risk language identifies the cause: customers may not expand "due to competing solutions" and "changes in their strategic IT priorities, budgets, costs."
Named competitors: Amazon DocumentDB and Azure Cosmos DB offer MongoDB-compatible APIs inside the cloud contract a customer has already signed. On a monthly-in-arrears basis with no commitment, a customer that consolidates a workload onto a hyperscaler-native equivalent simply stops consuming — there is no renewal event, no contract to breach, and no disclosure that would flag it until the net ARR expansion rate prints lower, and MongoDB publishes that figure once a quarter with no comparative (§4).
This is a slower and more insidious mechanism than SNOW's renewal right-sizing, because it has no renewal date at which it becomes visible.
Quantified. Net ARR expansion falling from 121% to 108% takes revenue growth to roughly 10%. With the multiple reverting to the post-2024-07 25th percentile of 9.2x: NTM revenue $2,863m × 9.2 = $26.3bn ÷ 80.5m = $327, +2% from spot — deceleration alone at a 25th-percentile multiple is roughly flat, which is what buying at the 21st percentile buys. The loss requires the multiple to fall further. - Expansion to 108%, multiple to the post-2024 minimum of 6.3x (printed April 2025): $2,863m × 6.3 = $18.0bn ÷ 80.5m = $224, −30% from spot. MDB traded at $245 in March 2026 and $210 in June 2025, so this is inside the realised range of the last fourteen months. - Expansion to 100% (no net expansion at all), multiple to 6.3x: $2,733m × 6.3 = $17.2bn ÷ 80.5m = $214, −33%. - A harder case — revenue flat and the multiple to 5.0x: $2,602m × 5.0 = $13.0bn ÷ 80.5m = $162, −50%. Net cash of $2,430.6m is $30.19 per share, 19% of that price.
Probability assigned: 25% for the −30% case, 10% for the −50% case.
Going-concern: no. $2,430.6m of net cash, zero debt (the 2026 convertibles were retired in 2024), $596.9m of TTM operating cash flow and $7.3m of TTM capital expenditure. MongoDB generated $201.6m of operating cash flow in the latest quarter alone while reporting a $24.8m operating loss. No going-concern case is argued.
The second-order risk, named: the RPO concentration. The $478m single-quarter step in RPO (Oct-25 $994.3m → Jan-26 $1,472.7m) is not attributed to a named customer or cohort, and 46% of the $1,458.6m base sits in the 13-to-36-month bucket. If that step represents a small number of large commitments, a single non-renewal in FY2028 would take RPO down by a fifth, and RPO is the only forward-looking contractual figure MongoDB publishes. That is not a reason to discount the name; it is the specific line to watch in each successive filing.