Marqeta — Blitz Index 43 (C)

Card-issuing infrastructure with blitz volume and a halved take rate; Block's 2028 renewal is the whole story.

Last re-scored: October 20, 2026.

Score breakdown

Think we scored it wrong? Score it yourself →

Radar axes (0–100): Moat 48.3 · Speed 60 · Creative 53.3.

Watch item

Gates: G1 ✓ · G2 ✓. Bar ≈ $7.0B (FY2026E net revenue × 10), computed on net revenue only — TPV is volume, not revenue.

Strengths

• TPV +31.4% (FY2025) and +31.8% (Q2 2026) — volume growth is durable
• Two straight profitable quarters and a 21% adjusted EBITDA margin — the profitability turn has started
• Gross margin steady at 70% with operating cash flow of $162.6M, up three years running
• Concentration is easing on paper: Block 68% → 41%, Sutton 76% → 58%

Concerns

• Take rate 0.304% → 0.146% and still falling in Q2 2026 — the floor is unconfirmed
• FY2026 net revenue guidance of +12–13% makes the volume/revenue gap explicit
• The Block renewal (June 2028) can reprice close to half of revenue in a single negotiation
• CEO and CFO both changed across 2025–2026 — execution continuity is an open file

Research notes

Thesis: a volume blitz that pays a falling unit price

Marqeta, Inc. is a card-issuing infrastructure company, founded in 2010 by Jason Gardner and listed on Nasdaq on 2021-06-09, whose business is best understood as two stories running in opposite directions at the same time. On one side, the volume story is a genuine blitz: total processing volume, or TPV, rose from $222.3 billion in fiscal 2023 to $291.1 billion, up 31.0 percent, and then to $382.5 billion in fiscal 2025, up 31.4 percent, with second-quarter 2026 TPV of $120.4 billion, up 31.8 percent year on year, all confirmed in Securities and Exchange Commission filings. On the other side, the unit-price story runs the other way: the take rate, the share of that volume converted into net revenue, fell from 0.304 percent in fiscal 2023 to 0.174 percent, then to 0.163 percent, and then to 0.146 percent in the second quarter of 2026. Volume grew by roughly 72 percent over two years, from $222.3 billion to $382.5 billion, while the take rate more than halved. That single tension explains almost everything about the score that follows.

The score file is explicit about its revenue definition, and the distinction matters for every number in this draft. The authoritative revenue measure is net revenue, which already reflects interchange fees net of customer revenue sharing. TPV of $382.5 billion is a processing-volume key performance indicator, not revenue, and it must not be mistaken for sales. The same net-revenue principle applied here was applied to comparable gross-to-net normalisations elsewhere in the scoring programme. Marqeta is classified as growth, with trailing-twelve-month net revenue of approximately $677 million and a market capitalisation of approximately $1.95 billion on a press-reported basis. The official BlitzLab v1.0 score at T, 2026-10-20, is 43 out of 100, grade C: speed 15 out of 25, moat 29 out of 60, creative destruction 8 out of 15, and deductions of 9 points. The total is 15 plus 29 plus 8 minus 9, which equals 43, a result 22 points below the 65-point line. There is no FLAG, because deductions of 9 sit below the 15-point FLAG line. On the radar conversion used in the score file, moat converts to 48.3, speed to 60.0, and creative destruction to 53.3 on a 100-point scale.

The gate verdict is a pass on both gates, so scoring proceeds. Gate 1, sustained growth or a second growth curve, passes on confirmed evidence: three consecutive years of volume growth above 30 percent, a recovery in net revenue after a sharp fall, a turn to quarterly profitability, and a second regional curve begun through the TransactPay acquisition, which closed in July 2025 and secured a direct issuing channel in the United Kingdom and the European Economic Area. Gate 2, a ten-year ceiling at least ten times current revenue, also passes, but on an estimate and with a condition attached, as discussed below. The company therefore clears the entry gates while landing well short of the 65-point line. That combination, passing the gates but scoring 43, is the shape of the whole assessment: a real platform business, already at scale in volume terms, whose economics are still being repriced by its largest customer.

Speed: volume at blitz pace, revenue at a slower pace

The speed section totals 15 out of 25, and its internal split shows where the pace is real and where it is constrained. Growth, item A1, scores 6 out of 10. The volume record alone would argue for a higher mark: TPV growth of 31.0 percent and then 31.4 percent across fiscal 2024 and 2025, followed by 31.8 percent year-on-year growth in the second quarter of 2026, is top-tier volume growth by any standard the score file applies. But the score is set by the gap between volume and revenue, not by volume alone. Net revenue fell 25.0 percent in fiscal 2024, from $676.2 million to $507.0 million, after the Block Amendment of August 2023, then recovered 23.3 percent to $624.9 million in fiscal 2025, and reached $176.0 million in the second quarter of 2026, up 17.0 percent year on year. First-half 2026 net revenue was $341.8 million, up 18.1 percent. Fiscal 2026 guidance, confirmed in a Securities and Exchange Commission filing, points to growth of only 12 to 13 percent, a clear deceleration from the volume pace. The score file benchmarks this against peers scored on revenue growth itself, noting that revenue growth a tier below the fastest growers supports 6 rather than a higher mark. The take-rate decline, from 0.304 percent to 0.146 percent, offsets more than half of the volume growth in revenue terms. Speed, in other words, is genuine in the pipes and muted in the income statement.

Capital velocity, item A2, scores 3 out of 5, and the reasoning is unusual: the evidence is not rapid fundraising but accumulated self-financing strength. Marqeta carries no long-term borrowings, so it is debt-free in the funded sense used by the score file. Liquidity at 2025-12-31, confirmed, was approximately $1,079.5 million, composed of cash of $709.4 million, restricted cash of $307.6 million, and short-term investments of $62.5 million. Operating cash flow was $162.6 million in fiscal 2025. In fiscal 2025 the company returned $391.4 million through share repurchases, meaning capital was being returned rather than raised. That is self-funded blitz fuel in reverse: the company is financing itself from operations rather than from new capital. The discount comes from the other side of the same picture. Market capitalisation of approximately $1.95 billion is press-reported, and the 1-for-4 reverse share split effective 2026-06-30, applied retroactively to all periods, is read in the score file as the market's judgement that the equity currency was weak. With no large-scale fundraising or investment episode demonstrated, the score stops at 3 rather than rising to 4.

Hiring and productivity, item A3, scores 3 out of 5. Headcount was 938 employees at 2025-12-31, confirmed. Net revenue per employee calculates to approximately $666,000, dividing $624.9 million by 938 people, while TPV per employee calculates to approximately $408 million. Platform leverage is therefore high in volume terms, but revenue per employee sits at roughly one third of the $1.81 million figure cited in the score file for a named peer. Stock-based compensation fell from $136.6 million in 2024 to $104.8 million in 2025, equal to 16.8 percent of net revenue, a declining trajectory that counts in the company's favour. Against that, the leadership transition was in progress through the scoring window: the chief executive resigned in February 2025, a permanent successor was appointed in September 2025, and the chief financial officer changed in February 2026. With no demonstrated hiring-velocity figure in the file, the score rests at 3.

Share velocity, item A4, also scores 3 out of 5. Volume growth above 30 percent suggests expanding share inside the card-issuing processing niche, but no quantified market-share figure is disclosed in the file. Against global card volume of $28.6 trillion, cited on a press-reported basis for 2023, TPV represents approximately 1.3 percent on a calculated basis. The only customer above 10 percent is Block, at 45 percent of fiscal 2025 net revenue, and there is no quantified demonstration of other large customers in the file. Entry into the United Kingdom and the European Economic Area through TransactPay counts as geographic extension, but it does not by itself convert a suggested niche-share gain into a measured market-share gain. The score file applies the same yardstick it applied to a peer whose brand momentum was not matched by confirmed absolute share, and the result is 3.

The ten-year ceiling: a pass that depends on price, not on market size

Gate 2 deserves separate attention because it is the most conditional pass in the file. The bar is set from an estimated fiscal 2026 net revenue figure of approximately $703 million, derived by applying the midpoint of the confirmed fiscal 2026 guidance range of 12 to 13 percent growth to fiscal 2025 net revenue of $624.9 million. Ten times that estimate gives a bar of approximately $7.03 billion, on an estimate basis. On the confirmed fiscal 2025 revenue base the bar would be $6.25 billion, and the file notes that the conclusion is the same on either base. TPV must not be used as the revenue base for this test: volume is not revenue, and the file explicitly rejects that definition.

The market-pool argument, on a press-reported basis, is large. Global card payment transaction value of $28.6 trillion in 2023 is projected by Allied Market Research, press-reported, to reach $56.4 trillion by 2033, a compound annual growth rate of 6.9 percent. Applying Marqeta's current take rate of 0.146 to 0.163 percent to that volume implies an issuing and processing revenue pool of approximately $90 billion in 2033, on an estimate basis, or roughly 13 times the bar. The market itself therefore clears the ceiling test by a wide margin. The file is careful, however, not to treat that arithmetic as comfort. At a fixed take rate of 0.16 percent, reaching the bar would require TPV of approximately $4.4 trillion, equal to approximately 8 percent of projected 2033 global card volume, against a current share of approximately 1.3 percent. If TPV continued to compound at its recent rate of above 31 percent for ten years, the arithmetic reaches approximately $5.7 trillion, above the bar, but that path assumes a sixfold expansion of share, and the file explicitly refuses to count it as headroom. The real bottleneck is not the size of the market but the unit price: the take rate has already fallen structurally since the Block Amendment, and the gate pass is explicitly conditioned on further structural decline stopping. If additional structural decline in the take rate is confirmed, the file names that as a trigger for re-examination. The distinction drawn against a peer that failed this gate is instructive: that peer's addressable spread pool was itself shrinking, with an optimistic ceiling below its bar, whereas Marqeta's volume pool is expanding at roughly 7 percent a year with a ceiling more than ten times its bar. The constraint is customer terms, above all the Block recontracting due in June 2028, not the ceiling of the market.

Moat: infrastructure quality without a classical network effect

The moat section totals 29 out of 60, the largest section and the one where the score most clearly reflects the character of the business. Network effects, item B1, score only 5 out of 18, and the file is explicit that there is no classical network effect. That is not presented as a missing feature to be excused; it is presented as the structural fact of a business-to-business infrastructure model. The moat that does exist consists of modern application-programming-interface issuing infrastructure, contrasted with legacy processors named in the file as FIS and Fiserv, together with real-time approval controls, programmable card logic, a bank-partner network, and compliance programme management. As programmes accumulate, risk and fraud data and issuing know-how accumulate as well, producing a weak data effect. The file places that in the lower part of its 5-to-9 anchor range for data and ecosystem effects. Because more end users do not raise the value of the service for other customers in the way a consumer network does, the upper part of the range is unreachable on this model. The moat is therefore real but bounded: execution quality in infrastructure, not a self-reinforcing user network.

Distribution scalability, item B2, scores 7 out of 18, at the upper end of the file's 5-to-9 anchor range for a model that depends on paid sales effort but can be extended repeatedly. The repeatable path is visible: direct sales, plus bank partners including Sutton among several, plus TransactPay's United Kingdom and European Economic Area electronic-money-institution licence and bank-identification-number sponsorship, which together give regional expansion a repeatable institutional route. This is also the first place where customer concentration is charged, following the precedent the file applies consistently. Block accounted for 45 percent of fiscal 2025 net revenue, and that single-customer dependence is treated here as a weakness in the distribution structure itself, just as comparable concentration was discounted in the distribution scores of peers cited in the file. The offset is a genuine de-concentration trajectory, confirmed: the Block share fell from 68 percent to 45 percent and then to 41 percent in the second quarter of 2026, while the Sutton settlement share fell from 76 percent to 58 percent. Both concentrations are easing, but neither has eased far enough to remove the discount.

Market size, item B3, scores 9 out of 12, reflecting the Gate 2 pass. The projected pool of approximately $90 billion against a bar of approximately $7.0 billion leaves large room on paper. The discount has two parts, both stated in the file: the market Marqeta can actually address, described as the modern application-programming-interface issuing slice, is much narrower than total card volume, and the falling take rate compresses the conversion of that pool into revenue. The score sits one point below a peer cited for a threefold order-book demonstration, a comparison the file makes explicitly.

Gross margin, item B4, scores 8 out of 12, and it is the strongest moat item in absolute terms. Gross margin rose from 49 percent in fiscal 2023 to 69 percent and then to 70 percent in fiscal 2025, and remained at 70 percent in the first half of 2026, on gross profit of $239.5 million. Gross profit in dollars rose 24.3 percent in fiscal 2025, from $351.8 million to $437.3 million. Adjusted earnings before interest, tax, depreciation and amortisation reached $109.6 million in fiscal 2025, a margin of 18 percent, rising to $36.8 million in the second quarter of 2026, a margin of 21 percent. Operating cash flow increased for three consecutive years, from $21.1 million to $58.2 million and then to $162.6 million. The evidence for software-like margins is therefore ample. The discount is structural and subtle: because the margin denominator is net revenue, the margin can look high even as the economics per unit of volume deteriorate, and gross profit per unit of TPV, approximately 0.11 percent, is falling. The margin rate is stable; the margin per unit of volume is not. Both facts are true at once, and the score of 8 reflects both.

Creative destruction: replacing legacy issuing, within limits

The creative-destruction section totals 8 out of 15, with each item in the middle of its range. Replaced order, item C1, scores 3 out of 5, at the file's anchor 3 for replacement under way but limited in scope. The replacement described is the shift from legacy-processor-dependent issuing, naming FIS and Fiserv, toward programmable application-programming-interface issuing, demonstrated in embedded card programmes from fintech-scale customers of the kind named in the file, including DoorDash, Klarna and Block. The direction is clear and the competitive picture in the annual report supports it, but the scale remains limited: TPV of approximately 1.3 percent of global card volume means the replacement, however real in its niche, has not yet reached the scale of a wholesale overturning. A full replacement at anchor 5, or a consumer-credit replacement of the kind scored at 4 for a named peer, would require a different breadth, and this business-to-business infrastructure layer does not yet show it. The score is therefore 3.

Irreversible behaviour change, item C2, also scores 3 out of 5, again at anchor 3 for change under way. The lock-in mechanism is concrete: moving a card programme to another bank or network requires reissuance, re-authentication and compliance re-approval, so a stack built on Marqeta's interfaces is difficult to unwind. The Block contract running to June 2028, with an automatic renewal structure, is presented as one face of that lock-in. The limit is who changes behaviour: the actor is an enterprise customer, a fintech building on the platform, rather than a mass consumer adopting a new standard. That difference in kind, rather than in degree, is why the score stops at the middle anchor instead of rising toward the anchor for mass-standard irreversibility. The score is 3.

Cost-structure change, item C3, scores 2 out of 5, the lowest creative-destruction item, and the reasoning is deliberately conservative. The premise of the business is that an application-programming-interface platform structurally lowers the time and cost of launching a card programme, but no dollar-measured collapse in unit cost is disclosed in the file, so no such collapse is claimed here. The take-rate decline from 0.30 percent toward 0.146 percent is explicitly not adopted as evidence of cost-structure change: the file reads it as the result of a large customer's bargaining power under the Block Amendment, not as efficiency passed back to customers. On a conservative principle, an unexplained price fall is not counted as a cost revolution. The score file applies the same yardstick to a peer scored at 2 for undisclosed measured evidence, and the result here is 2.

Deductions and risks: the cleanest balance sheet, the heaviest concentration

Deductions total 9 points, which is light by the standards of the scored fintech group in the file and is the reason no FLAG is raised. Debt, item D1, is only minus 1 out of a possible minus 10, and the file calls the debt trigger effectively absent. There are no long-term borrowings. Total liabilities of $763.1 million are operating in character, including revenue-sharing payables of $224.5 million and customer deposits of $306.9 million, confirmed. Liquidity of approximately $1,079.5 million and operating cash flow of $162.6 million sit alongside that operating liability base. The single point of deduction comes from the dilution axis: stock-based compensation of $104.8 million, equal to 16.8 percent of net revenue. Even there, the file notes the mitigant that fiscal 2025 share repurchases of $391.4 million exceeded stock-based compensation by 3.7 times, so the net share count was falling. Credit risk, the file adds, is borne by issuing banks and third parties as stated in the annual report, with no balance-sheet exposure of that kind for the company itself. The score file's own summary makes the point directly: among the scored fintech names, this is the lightest debt deduction, and the company's problem is not money but the quality of growth.

Revenue quality, item D2, is minus 4 out of a possible minus 10, and it is the largest single deduction. It has two components, and both are stated precisely. First, the consolidated securities class action known as Wai and Ford, filed in the Northern District of California in December 2024, alleged false statements concerning 2024 results, revenue and gross-profit outlook under the cited securities provision. On 2025-11-03 the parties reached an agreement in principle of $13.0 million, with a self-payment component of $5.0 million alongside directors-and-officers insurance, confirmed in Item 3 of the annual report. The file scores this one step more heavily than a peer scored at the allegation stage, because a settlement stage has been reached. Second, revenue quality is affected by single-customer concentration itself: 45 percent of fiscal 2025 net revenue from Block alone, charged here in revenue quality as well as in distribution, following the same dual-reflection precedent the file cites for a peer. The offsets are also recorded: no revenue restatement, no auditor change, and no audit-opinion issue, and the Block share is on a declining trajectory. The deduction is 4, not higher, because those offsets are real, but it is the largest deduction because a settled disclosure case plus near-half revenue concentration together go to the reliability of the revenue line itself.

Governance, item D3, is minus 3 out of a possible minus 5. Founder Jason Gardner owns 14,518,707 Class A shares, equal to 3.70 percent, plus 33,481,083 Class B shares, equal to 94.99 percent of that class, giving voting power of 46.92 percent, confirmed in the definitive proxy statement dated 2026-04-21. The officer and director group together owns 48.62 percent. The structure is dual-class, and founder voting power close to a majority limits the checking function of ordinary voting. The file places 46.9 percent at the lower end of its minus-3 range by reference to the anchor line drawn from peers at 74.3 percent and 58.8 percent. One mitigant is recorded: Gardner is not the chief executive, so the company operates under a professional chief-executive structure. That mitigant softens the governance concern without removing the structural fact of Class B concentration, and the deduction stays at 3.

Regulatory and social cost, item D4, is minus 1 out of a possible minus 5, the lightest form of regulatory deduction: exposure without a confirmed event. The structural exposures are dependence on issuing banks and card networks, so that a deterioration in a banking relationship would require volume to move, exposure to regulatory change including the Consumer Financial Protection Bureau, and Sutton Bank settlement concentration of 58 percent in the second quarter of 2026, on a declining trajectory. The annual report also names high attrition and turnover among senior executives as a risk. Against those exposures, the file records that no fines, regulatory measures or enforcement cases have been confirmed. Because the item is exposure rather than an actual event, the deduction is kept light at 1, and the file explicitly avoids double-counting: Block concentration, already charged in distribution and revenue quality, is not charged a third time here.

What to watch: the June 2028 negotiation and the price floor

The score file's own blitzscaling assessment places Marqeta in a settling phase after the first blitz act: volume has already reached scale at $382.5 billion, the centre of gravity of growth is moving from volume toward defending unit price and diversifying customers, and the turn to quarterly profitability is the entry condition for a second act spanning credit programmes, Europe and money movement. The file's judgement is that promotion above 65 would require either demonstration of a second large customer above 10 percent or confirmation that the take-rate decline has stopped. Neither condition is met in the file at T, and neither is assumed here.

The single largest risk named in the file is the June 2028 negotiation. The Block contract expiry is not presented as a routine renewal: the precedent of the August 2023 amendment, in which the take rate roughly halved, shows that recontracting reprices the unit economics rather than merely extending the term. The Block share has eased to 41 percent in the second quarter of 2026, which mitigates the exposure, but the absolute scale on a single negotiating table remains close to half of net revenue, and the file estimates the order of magnitude in the hundreds of millions of dollars. Whether diversification outruns the contract expiry is, in the file's phrase, the litmus test for whether the blitz continues. That test cannot be answered from the file at T, and it is recorded here as an open question rather than a prediction.

Three further markers follow from the score itself. First, the take-rate path: the file records decline continuing into the second quarter of 2026 at 0.146 percent, with no floor yet confirmed, so the trajectory of that single ratio is the most informative series to follow. Second, concentration: the Block share at 41 percent and the Sutton settlement share at 58 percent are both falling, and the pace of that fall relative to the June 2028 date is the measurable form of the diversification question. Third, profitability and self-financing: two consecutive profitable quarters, first-quarter net income of $7.8 million and second-quarter net income of $7.6 million, together with first-half net income of $15.4 million, adjusted earnings before interest, tax, depreciation and amortisation margin of 21 percent in the second quarter of 2026, and operating cash flow of $162.6 million in fiscal 2025, form the financial base from which any second act would be funded without new borrowing. The valuation note in the score file is not reflected in the score and is recorded here only as context, on a press-reported basis at 2026-10-16: a share price of $18.69, market capitalisation of approximately $1.95 billion, a price-to-sales ratio of approximately 2.9 times on trailing-twelve-month net revenue of approximately $677 million, and a price-to-earnings ratio of approximately 190, distorted by early-stage profitability, with a 52-week range of $14.80 to $22.28. Market capitalisation on that basis is approximately 28 percent of the Gate 2 bar of approximately $7.0 billion. The score itself, the file stresses, does not look at price.

One forward marker is dated after T and is recorded here only as a future marker with no anticipated outcome: third-quarter 2026 results were due on 2026-11-03, after the as-of date of 2026-10-20, and were therefore excluded from the score. No outcome for that release is anticipated or assumed in this draft. The assessment above stands entirely on filings and reporting available at T, and any later release would belong to a later reassessment, not to this one.

In sum, Marqeta at 43 out of 100, grade C, is a company that has already proved it can move volume at blitz pace, has built a high-margin, debt-free, self-financing platform, and has begun the harder work of diversifying away from the customer that made that volume possible. Its score is limited not by a shortage of cash or by a broken platform, but by a falling unit price, a concentrated revenue base, a settled disclosure case, and a governance structure that concentrates voting power. The file's closing diagnosis captures the balance precisely: the risk is not blitz failure but settlement as a low-unit-price volume business. Whether the second act, in credit, in Europe, or in money movement, lifts the unit economics back toward the volume story is the question the score leaves open, and it is the question on which any future reassessment will turn.

More Platforms companies

Other Platforms names on the Blitz Index, highest score first: Alphabet (83/A), Meta (78/B), Airbnb (76/B).

Next earnings

Q4 2026 · Expected

Coverage

Marqeta, Inc. · Q3 2026 due Nov 3, 2026 (AMC) · Score 43 · Grade C

Embed this score

Copy the HTML below to show this score on your site. The badge links back to this page and updates automatically when the score changes.

<a href="https://blitzlab.app/companies/mq/"><img src="https://blitzlab.app/badge/mq.svg" alt="Marqeta Blitz Index 43/C — BlitzLab"></a> Copy HTML

Preview:

BlitzLab scores businesses against a scholar’s standard — not investment advice. No price targets, no buy/sell recommendations. Figures are from SEC filings (XBRL) as labeled in the source scorecard.