Every vendor will tell you their software is "affordable." Almost no one publishes what their customers actually spend in total — across the core ledger, the bolted-on tools, the middleware, the per-user fees, and the implementation. We surveyed 500 small and mid-sized businesses in Q1 2026 and asked them, line by line, what their finance stack costs. This report is what we found.
Some of it is obvious. Some of it isn't. The most striking finding: the median 50-to-200 employee company spends more on the tools attached to their accounting system than on the accounting system itself. The integration tax is real, and it's growing.
Methodology
We surveyed 500 SMBs between January and March 2026. Eligibility: between 1 and 500 employees, US- or Canada-headquartered, currently using one or more accounting software products. Respondents were finance leaders (CFO, controller, VP Finance, head of accounting, owner-operators in the smallest tier) and were compensated $50 USD for completing the survey. Sample composition:
- By size: 41% under 25 employees, 32% 25–100, 21% 100–250, 6% 250–500.
- By geography: 78% US, 22% Canada.
- By industry: 27% professional services, 19% SaaS/technology, 14% retail/e-commerce, 11% manufacturing, 10% nonprofit, 8% healthcare, 6% construction/trades, 5% other.
We asked respondents to itemize every recurring software cost in their finance stack — accounting platform, payroll, expense management, AP automation, AR/invoicing, billing, FP&A, close management, document capture, banking middleware. Costs are reported in USD per month, normalized from annual contracts where applicable. Where companies operate multiple entities or geographies, we asked for total combined spend.
One caveat. Self-reported spend has known biases — respondents undercount add-ons and forget per-user surcharges. We tried to mitigate this with line-item prompts, but the real numbers are almost certainly modestly higher than what's reported here. Treat these as conservative.
Five things we learned
1. The median 50-to-200 employee company spends $1,820/mo on accounting software — but only 38% of that on the core ledger
The headline number for mid-market SMBs (50–200 employees) is a median total finance software spend of $1,820 per month, or about $21,800 per year. But just $695 of that goes to the actual accounting platform. The remaining $1,125 is split across expense management ($310), payroll integration and add-ons ($240), AP automation ($175), document capture ($95), FP&A or budgeting ($165), close management ($85), and miscellaneous middleware ($55). The accounting platform is the foundation, but it's the minority of the bill.
Imagine a stacked bar chart here: the bottom 38% is "core accounting," the next 17% is "expense management," and the rest is a mosaic of integrations. The point lands at a glance.
2. The "integration tax" is now ~23% of total finance software spend
Across all respondents, we calculated the share of spend that goes to tools whose sole purpose is to move data between other tools — bank feed aggregators, data sync platforms, AP/AR middleware, OCR-only tools, sync-to-GL adapters, and the like. The median was 23%. For companies with three or more entities, this jumps to 31%. That's almost a third of finance software budget spent connecting systems that should already be connected.
This is the most actionable finding in the report. Every dollar in the integration tax bucket is a dollar that doesn't produce finance insight — it just keeps the lights on between systems.
3. AI add-ons are the fastest-growing category — 4.2x year over year
In Q1 2025, our same-cohort respondents spent a median of $32/mo on AI-related add-ons (categorization, anomaly detection, AI close assistants, AI-coded journal entries). In Q1 2026, that's $134/mo — a 4.2x increase. Two-thirds of mid-market respondents now pay for at least one explicitly AI-branded finance tool, up from less than a third a year ago.
The interesting wrinkle: respondents who have moved to AI-native platforms (where AI is embedded, not bolted on) reported lower total AI spend — they're not paying twice. That's an early but real signal that the AI-as-add-on category is a transitional one.
4. Manual reconciliation and OCR-only tools are in decline
Two categories shrank meaningfully: dedicated manual reconciliation tools dropped from a median $58/mo to $19/mo (down 67%), and OCR-only document capture (Dext, Hubdoc, point-tool variants) dropped from $48/mo to $34/mo (down 29%). The why is intuitive — both functions are being absorbed by AI-native ledgers that capture, code, and reconcile in one workflow.
If you sell OCR or rec tools as your main product, the data is not flattering. If you're a buyer, the question is whether you can consolidate these line items into your core platform.
5. The "QuickBooks plus eight things" stack costs more than NetSuite at ~150 employees
This one surprised even us. Among 100–250 employee respondents using QuickBooks Online as the core ledger, the median total finance software spend (QBO + Bill.com + Ramp + Gusto + a sync tool + a budgeting tool + close software + a couple of others) was $2,640/mo. Among similarly-sized respondents on NetSuite, the median was $3,150/mo — only 19% more, before counting the SI implementation amortized cost.
The implication: the cost gap between "scrappy SMB stack" and "real mid-market platform" has narrowed dramatically. At a certain size, the question isn't "can we afford to upgrade" — it's "are we already paying upgrade-tier money without getting upgrade-tier benefits."
Spend by company size
Median total finance software spend per month, USD:
| Company size | Core accounting | Add-ons & integrations | Total median spend |
|---|---|---|---|
| 1–10 employees | $45 | $85 | $130 |
| 11–25 | $135 | $220 | $355 |
| 26–50 | $295 | $510 | $805 |
| 51–100 | $520 | $890 | $1,410 |
| 101–200 | $845 | $1,310 | $2,155 |
| 201–500 | $1,720 | $2,485 | $4,205 |
The non-linearity is the story. Going from 25 to 100 employees roughly quadruples your finance software bill — well faster than headcount growth itself. That's the cost of bolting on capability instead of upgrading the platform.
Spend by industry
Median total finance software spend per month, USD, for companies in the 50–200 employee band:
| Industry | Median monthly spend | Top driver |
|---|---|---|
| SaaS / Technology | $2,485 | Billing & revenue recognition add-ons |
| Healthcare | $2,210 | Compliance, expense, multi-entity |
| Manufacturing | $2,095 | Inventory, AP automation |
| Professional services | $1,640 | Project profitability, time tracking |
| Nonprofit | $1,540 | Fund accounting, grants management |
| Retail / E-commerce | $1,830 | Multi-channel reconciliation, inventory |
| Construction / Trades | $1,920 | Job costing, AP, payroll |
SaaS and healthcare lead, both for industry-specific reasons. SaaS pays a heavy revenue recognition and billing tax (Stripe + Maxio + a billing engine + close software). Healthcare pays for compliance and multi-entity overhead. Professional services, despite often being a more sophisticated finance function, tends to spend less per dollar of revenue — services workflows have a smaller add-on ecosystem.
What's growing, what's shrinking
Categories ranked by 12-month change in median spend across the full sample:
| Category | 12-month change | Direction |
|---|---|---|
| AI add-ons | +319% | Growing fast |
| AP automation | +34% | Growing |
| Expense management | +22% | Growing |
| FP&A / planning | +18% | Growing |
| Core accounting platform | +9% | Stable (price increases) |
| Payroll integrations | +4% | Stable |
| OCR-only document capture | -29% | Shrinking |
| Manual reconciliation tools | -67% | Shrinking fast |
The pattern is clear: budget is moving from "human-replacement single-purpose tools" (OCR, manual rec) toward "intelligent assistance" (AI-coded journals, anomaly detection, AI-close support). Whether you're a vendor or a buyer, that's the through-line of 2026.
The integration tax, in detail
We asked a sub-question: how much of your finance software bill exists only to move data between other tools — middleware, sync services, bank feed aggregators, point integrations? The median across all respondents was 23% of total spend. Distribution:
- Bottom quartile: 8% of spend on integration glue.
- Median: 23%.
- Top quartile: 38%.
- Top decile (multi-entity, multi-tool): 49% of total finance software spend is moving data between systems.
Half of your finance software budget moving data is a sign your platform is the wrong shape for your business, not that you've optimized your stack. This is exactly the math that makes integrated, MCP-native platforms attractive — replacing five integrations with one shared data model is a hard dollar saving, not just a workflow nicety.
Recommendations
Based on what we saw in the data, three concrete suggestions:
- Audit your integration tax annually. Tag every line item in your finance stack as either "produces insight" or "moves data." If "moves data" is over 25% of your spend, you've outgrown your platform shape.
- Be skeptical of AI add-ons priced as separate SKUs. The AI-as-add-on market is a transitional one. AI-native platforms are absorbing those features into the core product. Paying separately for what's becoming a baseline feature is a near-term overpay.
- At 100 employees, run the full math on staying vs. upgrading. The cost gap between a sprawling SMB stack and a real mid-market platform has compressed. The decision is no longer purely financial — it's "what's worth the migration effort."
Built to compress the integration tax
VeloLedger replaces the GL plus expense plus AP plus close software plus the sync tools between them — with one MCP-native data model. See what your stack could look like.
See a demo