Revenue Architecture for Expense Management Software — The Complete Operator Guide in 2027

The two revenue models operators actually choose between

Every expense management software company built after 2020 faces a fork that does not exist in most other B2B categories: you can charge for the software, or you can charge for the money that moves through it. These are not variations on a pricing page. They are structurally different businesses with different cost of goods, different sales motions, different comp plans, and different regulatory exposure.

Model A — subscription PUPM. The classic SaaS shape. You license seats on a per-active-user-per-month basis, recognize revenue ratably, and expand through seat growth plus module attach. SAP Concur is the archetype at roughly $2.1B in segment revenue across a customer base in the tens of thousands. Expensify runs the same model at the small end — roughly $135M in revenue but spread across hundreds of thousands of customers, which tells you the average contract is tiny and the motion is almost entirely self-serve. Coupa's expense segment sits inside a broader spend-management suite. The economics here are familiar: 75–85% gross margin, revenue that is highly predictable, churn that shows up on a renewal date you can see 12 months out.

Model B — interchange-funded. You issue a corporate card (through Visa or Mastercard rails, typically via an issuing partner like Stripe Issuing or Marqeta), capture a share of the interchange on every dollar of customer spend, and give the software away or price it near zero. Ramp and Brex are the reference implementations — both grew to substantial ARR (Ramp publicly discussed as $700M+, Brex in the $300M+ range at their last disclosed marks) with software that competitors charge $9–22 PUPM for offered free. Gross margin on interchange is lower than SaaS but the revenue scales with the customer's spend, not their headcount, which is a fundamentally better expansion curve.

The blend is what actually wins. Pure Model A vendors watch their mid-market pricing get compressed by competitors who charge nothing. Pure Model B vendors carry regulatory risk (see the Durbin expansion discussion below) and revenue that contracts hard in a downturn when travel and discretionary spend collapse. The operators building durable revenue architecture in 2027 run both: a subscription floor that survives a spend contraction, plus interchange upside that compounds when the customer grows.

The practical consequence for revenue architecture is that your quota, your forecast, and your retention math all have two axes instead of one. A rep who closes a $180K subscription deal with zero card attach has produced a materially worse customer than a rep who closes $60K in subscription with $40M in annual card volume routed through your program. If your comp plan cannot tell those two deals apart, your revenue architecture is broken regardless of how good the rest of it looks.

Where the third model lives. Navan represents a variant worth naming separately: travel-plus-expense bundling, where booking fees and travel supplier economics join subscription and interchange as a third revenue line. This matters competitively because it lets a bundled vendor undercut a T&E-only vendor on the expense line while making it up on travel — the same squeeze play interchange runs against pure SaaS, executed from a different direction.

How to decide which model your company should run

The choice is not preference. It is determined by three inputs: the segment you sell into, the sophistication your product actually has, and whether you can get an issuing partnership on economics that work.

Segment determines feasibility. Interchange-funded models work beautifully at SMB and lower mid-market, where the buyer is a founder or controller who wants one tool, has no incumbent, and will happily route spend through a new card. They work poorly at 5,000+ employee enterprises, where the company already has a banking relationship with a global bank, a negotiated commercial card program, and a treasury team that will not move card volume to a software vendor. If your ICP is Fortune 1000 global T&E, interchange is at best a partial revenue line — you sell subscription and you sell it on capability.

Product sophistication determines defensibility. The free-software squeeze only works where the free product is good enough. Multi-currency handling, global VAT and tax reclaim, IFRS and local statutory compliance, per-entity chart-of-accounts mapping, and integration depth with SAP or Oracle ERP are genuinely hard. A vendor with that depth can hold $22–55 PUPM enterprise pricing against a competitor giving software away, because the competitor's product cannot do the job. A vendor without that depth competing at mid-market against free software is in a price war it will lose.

Issuing economics determine whether Model B is even available. You do not simply decide to earn interchange. You need an issuing partner, a BIN sponsorship arrangement, credit underwriting capability (or a charge-card structure that avoids it), and a negotiated share of interchange that leaves you enough margin after network fees, rewards, fraud losses, and servicing costs. That is a 9–18 month build with real balance-sheet implications. A Series A company deciding to "add a card" is signing up for a fintech operating burden, not a product feature.

Read the decision tree as a sequence, not a menu. Segment first, because it constrains everything downstream. Issuing capability second, because it is a hard yes/no that takes quarters to change. Product depth third, because it is the one input you can actually invest in over a planning cycle. Companies that run this in the wrong order — deciding on a card program before knowing whether their ICP will route spend to it — burn 12–18 months and a fintech compliance team discovering the answer.

The concrete numbers behind each model

Here is where the two models diverge in the operating plan, tier by tier.

Tier 1 — Strategic Enterprise (5,000+ employees, global T&E plus card program). Roughly four thousand US enterprises fit this profile. ACV bands run $185K to $2.8M depending on seat count and module attach; a full T&E + corporate card + AP automation + travel deployment at 5,000+ employees typically lands in the $385K–$2.4M range before any interchange contribution. Pricing is $22–55 PUPM. Sales cycle is 3–7 months. Coverage requirement is 3.5x on a rolling-three-quarter basis, 2.8x in-quarter. Win rate floor is 26% — Gartner's travel-and-expense coverage puts category win rates in a 22–48% band, and an enterprise rep sitting under 24% is a coaching trigger, not a bad-luck story.

Tier 2 — Mid-Market (200–5,000 employees). Roughly 78,000 US firms. ACV $18K–$185K. Pricing $9–22 PUPM, or free-with-card if you are running the interchange model. This is where the price war lives: mid-market PUPM compressed meaningfully over 2024–2026 as free-software competitors took share. Sales cycle is 2–6 weeks. Coverage 3x rolling-two-quarter. Win rate floor 38%.

Tier 3 — SMB (under 200 employees). Roughly 1.4 million firms. ACV $1.5K–$18K. Pricing $4–9 PUPM. Sales cycle 1–3 weeks, heavily product-led. Coverage 2.5x rolling-one-quarter. Win rate floor 50%, which sounds high until you remember the funnel that feeds it is self-serve trials, not qualified enterprise opportunities.

Funnel conversion, stage by stage. MQL-to-SQL runs about 28% at Tier 1, 38% at Tier 2, 52% at Tier 3 — the enterprise number is low because the qualifying contact must be a CFO or Controller, not an office manager. SQL-to-discovery: 58/65/75. Discovery to demo-or-POC: 42/52/60. POC to procurement: 52/60/68. Procurement to closed-won: 26/38/50. Multiply it through and total funnel conversion is roughly 0.9% Tier 1, 2.8% Tier 2, 5.9% Tier 3. Those are the numbers your demand gen plan has to work backward from.

Compensation, by role. Strategic Enterprise AE: $275–315K OTE at a 50/50 split, carrying $1.0–1.4M. Mid-Market Territory AE: $175–205K OTE at 60/40, carrying $550–700K. SMB Inside AE: $115–135K OTE at 65/35, carrying $375–475K. SDR/BDR: $80–100K OTE at 70/30 against 12–18 SQLs monthly. Strategic CSM: $155–185K OTE at 70/30, gated on 130% NRR and 90% GRR. Mid-Market CSM: $115–135K at 85/15. Implementation Manager: $135–165K at 80/20.

The role that only exists in this category. A Card Spend Specialist overlay at $165–195K OTE, 60/40, carrying a card-spend-volume quota rather than an ACV quota. Their job is converting subscription customers into interchange-paying customers — driving card attach toward 65%+ of the base. At card-led vendors, roughly 25% of AE variable comp is tied to first-year card spend volume rather than subscription bookings, which is the single most important comp-design decision in the category. Alongside them sits a VP of Bank Partnerships ($235–285K OTE, 75/25) who owns issuing relationships and interchange-share negotiation. At a card-led vendor, that function influences 30–50% of total company revenue and reports into the CRO with a hard dotted line to the CFO, because interchange revenue recognition is materially more complex than ratable subscription revenue.

Accelerators and ramp. Standard structure is 1.5x on incremental attainment to 100%, 2.5x above 125%, with a decelerator to 50% below 70% attainment. Enterprise AE ramp is 30% of quota in Q1, 65% in Q2, 100% by Q3 — a six-month curve. Mid-Market ramps at 50% then 100% over four months. SMB ramps 75% then 100% over three months.

Retention economics, which is where the models separate hardest. Gross revenue retention in the category runs 88–94% best-in-class. Net revenue retention targets 122–135%. But the composition matters enormously: a customer with card attach expands at 150%+ NRR because their card spend grows with their business, while a subscription-only customer with flat headcount compresses toward 95%. The math on a healthy blended NRR is roughly 92% GRR, plus 3–5% from employee growth driving seat expansion, plus 18–35% from card spend growth, plus 8–14% from AP automation attach. Strip the card line out and 130% NRR is not reachable through seats alone.

Implementation and sequencing — building the engine in the right order

The sequencing failure that kills expense management revenue orgs is hiring the enterprise team before the product can hold enterprise price, or standing up a card program before anyone has proven customers will route spend to it. Here is the order that works.

$0–5M ARR. Founder-led sales plus one solutions engineer. Do not hire a sales team. The job at this stage is finding the twenty customers whose problem your product actually solves better than the incumbent, and learning whether they will move card spend. If you intend to run an interchange model, this is when you validate the attach rate — not after you have committed to an issuing partnership.

$5–20M ARR. Two to four inside AEs, first SDR, first CSM, and — critically — the first Card Spend Specialist. That last hire is early on purpose. Card attach is a learned motion, and the vendor who figures out how to get a controller to move their commercial card program in month three of the relationship has a structurally better business than one who tries to bolt it on at $50M. All of this reports to a VP Sales.

$20–60M ARR. First Strategic Enterprise AE, second SE, first Strategic CSM, a RevOps lead, and a VP of Bank Partnerships. The enterprise hire should follow, not precede, the first Tier 1 closed-won — if the founder cannot close a 5,000-employee account, a new AE will not either. RevOps moves under the CRO here.

$60–200M ARR. RVP Enterprise, RVP Mid-Market, Director of CS, VP Card Solutions, VP Travel Partnerships. This is the multi-product scale point where the org stops being one motion and becomes several.

$200M+ ARR. Director of RevOps Analytics (specifically for card-volume cohort modeling), VP Product Marketing, vertical leads, and a VP Strategic Alliances covering the ERP and accounting ecosystem — NetSuite, QuickBooks, Workday, SAP. Rule of thumb on RevOps headcount: roughly one FTE per $20M ARR, with at least three analysts dedicated to card-volume cohort modeling and interchange revenue forecasting once you cross $300M.

Forecast methodology has to be finance-calendar aware. Expense management buyers are finance teams, and finance teams buy on their own calendar. Q4 budget reload drives roughly 32% of annual bookings; the Q1 January-go-live surge drives another 26%. Nearly six in ten deals land in two quarters. Build a three-bucket model: Commit is 80%+ probability with CFO and Controller sign-off plus a scheduled bank-card setup date; Best Case is 50–79% with demo complete and card program scoped; Pipegen is 25–49% with qualified discovery only. Reconcile Monday/Wednesday/Friday weekly, with monthly NRR and card-spend cohort review. AI forecasting tools help here, but only if you feed them category-specific signals: incumbent renewal dates, card-spend volume trend (growth or contraction), and travel volume rebound.

Expansion comp triggers should be explicit, not discretionary. Card spend expansion pays the Card Specialist a SPIFF at roughly 12% of incremental annual spend. AP automation attach is AE-led with the CSM attached at 30% credit. Travel integration attach follows the same structure. Seat true-up is CSM-owned at 25% of the seat uplift. Without named triggers, expansion revenue gets claimed by whoever argues loudest in the QBR, and your best expansion motion quietly stops happening.

Renewal risk scoring needs category-specific signals. CFO turnover within nine months of renewal is a red flag — the new CFO reviews every vendor. Card spend dropping more than 20% in 60 days is yellow, and it is usually not a product problem: it means the customer's business is contracting, which predicts a seat reduction two quarters out. And regulatory movement on interchange is a sector-wide yellow that hits every card-led customer simultaneously.

The failure modes that break this architecture

Incumbent enterprise lock-in. SAP Concur holds well over 30% of enterprise share with ERP integration depth accumulated over two decades. You do not displace it on feature parity. The realistic wedge is implementation speed (90 days versus 9 months), cloud-native user experience that employees will actually adopt, and AI-driven receipt and policy processing. Plan enterprise displacement cycles at 5–7 months minimum and expect to lose more than you win.

The free-software squeeze. Interchange-funded competitors giving away software compressed mid-market PUPM roughly 22% across 2024–2026. If you sell subscription-only into mid-market, this is an existential trend, not a competitive annoyance. The two viable responses are moving upmarket into complexity the free products cannot handle, or building your own card program. "Explaining our value better" is not a third option.

Interchange regulatory risk. Proposed expansion of interchange regulation to credit cards has been under legislative discussion in the US. If enacted in a broad form, it could compress interchange revenue meaningfully — the operator planning assumption is a 30–50% haircut in a severe scenario. Any company where interchange is more than a third of revenue needs a diversification plan on the shelf: subscription floor, AP automation, travel, procurement adjacency.

Spend contraction in a downturn. Card spend, particularly the travel component, drops sharply in recession quarters — an 18–28% decline is a reasonable planning band. Interchange revenue falls with it, immediately, with no contractual floor. This is the argument for multi-year subscription commitments even at card-led vendors: they are not a revenue maximizer, they are a downside floor.

The travel bundle squeeze. A competitor bundling travel booking with expense can price the expense line aggressively and recover margin on travel. T&E-only vendors get squeezed from a direction their pricing model does not defend. The counter is depth in the accounts-payable and procurement direction rather than the travel direction — different adjacency, same defensive logic.

Operating cadence that keeps this visible. Weekly: strategic pipeline review, RevOps roll-up, card-spend cohort review, CS escalation. Monthly: NRR and GRR cohort review, card-spend trend analysis, AP attach cohort. Quarterly: territory rebalance, comp plan retrospective, issuing bank partnership review, travel partnership review. Annually: ICP refresh against regulatory shifts, full comp plan rebuild. The card-spend cohort review is the meeting most companies skip and the one that would have caught their NRR compression two quarters early.

Related questions

Should a subscription-only expense vendor add a corporate card?

Only if your ICP will actually route spend. SMB and lower mid-market will; large enterprises with existing commercial card programs and treasury relationships generally will not. Validate attach intent with twenty customers before committing to a 9–18 month issuing build and the fintech compliance burden that comes with it.

How do you comp a rep on card volume without breaking the plan?

Run dual quotas — subscription ACV plus first-year card spend — with roughly 25% of variable comp on the volume line. Pay card spend on realized volume at 90 days, not projected, or reps will book programs that never activate.

What NRR should an expense management vendor target?

122–135% net revenue retention against 88–94% gross retention. The composition matters more than the headline: card spend growth contributes 18–35 points, AP attach another 8–14, seat growth only 3–5. Subscription-only vendors realistically top out near 105–110%.

Why do so many deals land in Q4 and Q1?

Finance teams buy on the budget calendar. Q4 budget reload drives roughly 32% of bookings and the January go-live surge drives another 26%. Build capacity plans and hiring around that concentration rather than assuming even quarterly distribution.

When should the first Card Spend Specialist be hired?

At $5–20M ARR, alongside the first CSM — earlier than most companies think. Card attach is a learned motion, and vendors who master it before scale carry a structurally better expansion curve than those bolting it on at $50M+.

FAQ

What is the typical sales cycle for enterprise expense management in 2027?

Three to seven months at Tier 1 enterprise (5,000+ employees, global T&E plus card program), two to six weeks at mid-market, and one to three weeks at SMB. The enterprise cycle stretches because it requires CFO and Controller sign-off, procurement review, security review, and — where a card program is involved — treasury and banking coordination that no software vendor controls.

How does the interchange revenue model change compensation design?

It adds a second quota axis. Card-led vendors put roughly 25% of AE variable compensation on first-year card spend volume rather than subscription bookings, and add a dedicated Card Spend Specialist overlay at $165–195K OTE on a 60/40 split carrying a pure volume quota. Without that, reps optimize for ACV and ignore the revenue line that actually compounds.

Can a mid-market vendor compete head-on against free interchange-funded software?

Only in two situations: customers who explicitly refuse to bundle their card program with a software vendor (more common in regulated industries and companies with existing banking relationships), or where your product handles complexity — multi-currency, statutory compliance, multi-entity accounting — that the free products do not. Competing on positioning alone against a $0 price point does not work.

What RevOps headcount does a $300M expense vendor need?

Roughly one RevOps FTE per $20M ARR, so 15 or so at $300M. At least three of those should be analysts focused specifically on card-volume cohort modeling and interchange revenue forecasting, because that revenue line does not behave like subscription and cannot be forecast with subscription tooling.

How should a card-led vendor plan for interchange regulatory risk?

Model a 30–50% interchange compression as a downside scenario and check what your P&L looks like under it. If the answer is unviable, you need diversification now — subscription floor pricing, AP automation, travel, or procurement adjacency — rather than after legislation moves. Multi-year subscription commitments also provide a contractual floor that interchange revenue lacks.

What is the single biggest structural mistake in this category?

Treating card attach as a post-sale upsell rather than a first-year revenue commitment. Attach rates achieved in the first 90 days after go-live run dramatically higher than attach attempted at renewal, because the implementation window is when the customer is actively rewiring their spend process. Vendors who wait until the QBR to raise card programs never reach 65% attach.

Sources

• https://www.gartner.com/en/information-technology/glossary/travel-and-expense-management
• https://www.sap.com/investors/en/reports.html
• https://investors.expensify.com/
• https://www.federalreserve.gov/paymentsystems/regii-about.htm
• https://www.congress.gov/bill/118th-congress/senate-bill/1838
• https://usa.visa.com/support/consumer/visa-rules.html
• https://www.mastercard.us/en-us/business/overview/support/rules.html
• https://stripe.com/issuing
• https://www.marqeta.com/platform
• https://www.forrester.com/research/

Related on PULSE

• [Revenue Architecture for Trading + Order Management Systems (OMS) Software in 2027 — The Complete Operator Guide](/knowledge/ra0100)
• [Revenue Architecture for Risk Management / GRC Software in 2027 — The Complete Operator Guide](/knowledge/ra0093)
• [Revenue Architecture for Court + Case Management Software in 2027 — The Complete Operator Guide](/knowledge/ra0089)
• [Revenue Architecture for Fleet Management Software in 2027 — The Complete Operator Guide](/knowledge/ra0076)
• [Revenue Architecture for TMS (Transportation Management Software) in 2027 — The Complete Operator Guide](/knowledge/ra0074)
• [Revenue Architecture for WMS (Warehouse Management Software) in 2027 — The Complete Operator Guide](/knowledge/ra0073)