Small Team Expense Reports
Without the $200 Per-User Premium

Per-User Pricing: The Math That Breaks at Small Scale

The per-user pricing model is the SaaS industry's default, and it wasn't built for expense reports. It was built for CRMs and project management tools — software where every user logs in daily, generates value proportional to their time in the system, and the vendor's server cost scales with active sessions. Expense reports don't work that way. An employee submits once or twice a month, and the submission takes five minutes. The rest of the month, that seat generates zero server load. But the vendor bills for it as though it did.

For a 10-person team, the math looks like this before anyone touches a receipt:

The problem with these numbers isn't that they're expensive in absolute terms — $100 a month for ten people automating expense processing sounds reasonable if all ten people are submitting expenses. The problem is that at small scale, they never are. A Spendesk analysis of 332,000 business subscriptions found that 66% of SaaS subscriptions go unused in a given month. Expense management tools are disproportionately affected: they're provisioned to every employee in the company, but used by the subset that actually travels or makes reimbursable purchases. In a ten-person design studio, that subset might be four people. In a fifteen-person construction office, maybe six. The other seats are a monthly recurring charge for software that nobody opens.

Nor is this a matter of picking a cheaper per-user plan. Zoho Expense's $4/user brings a 10-person team to $40 per month, which is the most affordable per-user option in the market — but it still charges $4 for the developer who submitted one Uber receipt in March and nothing since. The pricing model itself is the problem, not the rate.

What Small-Team Expense Volume Actually Looks Like

Industry benchmarks for expense report volume are skewed upward by enterprise data. The Global Business Travel Association (GBTA) Foundation found that the average expense report costs $58 to process and takes 20 minutes from submission to reimbursement — with 19% of reports containing errors that add another 18 minutes and $52 each to correct. But those averages pool Fortune 500 companies with hundred-person sales teams alongside five-person consultancies. At small-team scale, the volume is an order of magnitude lower.

What does actual expense volume look like for a team of five to twenty people?

The gap between team size and active submitters isn't random. Salespeople and field staff submit monthly because travel and client meals are built into their role. Engineers and designers submit once every few months, when they attend a conference or buy a software license on a personal card. Administrative staff and interns rarely submit at all — their purchases go through the company card. A per-user pricing model charges for all three groups equally, even though their usage patterns are nothing alike.

There's a second dimension to the mismatch: volume is lumpy, not steady. A nonprofit with field staff might process 20 expense reports in June — the end of the program cycle when everyone clears their receipts — and five in July. An engineering consultancy processes 15 reports in March when the annual conference lands, and three in February. Subscription pricing assumes steady monthly usage. Small-team expense reality is bursty. When the monthly fee doesn't vary with volume, slow months become overpayments, and busy months leave you outside the included quota hunting for overage pricing.

When Expense Management Software and Expense Extraction Are Different Products

Expense management platforms do three things. They collect expenses — employees snap photos of receipts, fill in amounts and categories, and attach business-purpose notes. They manage approvals — reports route through managers, finance, and compliance, with policy rules flagging out-of-policy purchases. They handle reimbursement — approved amounts land in the employee's bank account and sync to the general ledger. This is the full lifecycle. It's why Expensify and Concur call themselves "expense management" and not "expense extraction."

Expense extraction is one slice of that lifecycle: read the receipt or expense report, identify the relevant fields — date, vendor, amount, category, business purpose, payment method — and output them in a structured format. What happens after extraction — whether the data flows into an approval queue, into QuickBooks, or into a spreadsheet that feeds a manual check-writing process — is downstream. It's a separate layer.

This distinction matters because most small teams don't need the full lifecycle. They need the extraction slice. A 10-person creative agency doesn't have a multi-step approval workflow — the founder reviews expenses once a month, approves the totals, and cuts reimbursements. They don't need policy enforcement — the policy is "if it's for a client, it's reimbursable, if it's not, ask first." They don't need ERP integration — they use QuickBooks or Xero, and their bookkeeper enters a handful of journal entries. What they need is the data out of the receipts — structured, accurate, and available without two hours of manual transcription every month.

Buying a full expense management suite for that need is like subscribing to an inventory management platform because you need to scan ten barcodes a week. The platform does what you need, but you're paying for everything else it does too — for everyone in the company, whether they use it or not.

The point is not that expense management suites are bad products — for teams that need audit trails, approval routing, and reimbursement automation, the full lifecycle is worth the per-user fee. The point is that most small teams don't need that lifecycle. They need the data. Paying for the lifecycle to get the data is the pricing trap that this article exists to help you avoid.

Collection Link: How Shared-Pool Access Replaces Per-User Login

The per-user pricing model is rooted in a technical assumption: every person who submits expenses needs a user account — with a login, a password, a profile, and permissions. If you want ten people to upload receipts, the platform reasons, you need ten accounts. That assumption made sense when the only way to submit was through a logged-in dashboard. It no longer holds.

A Collection Link is a shareable URL — it looks like any other web link — that, when opened, presents a file upload page. The recipient enters a short verification code, drops their receipt files into the upload area, and closes the browser tab. They don't create an account. They don't remember a password. They don't appear on a user roster. The files land in the administrator's processing queue automatically, tagged and ready for extraction.

This mechanism decouples upload access from user accounts. Instead of provisioning ten seats for ten people who might submit expenses, you generate one Collection Link and share it with the team — through Slack, email, or a pinned message. Whoever has the link can upload. Whoever doesn't have expenses to submit simply never opens it. The cost model shifts from "per person who could submit" to "per document that actually gets submitted."

In practice, the difference looks like this for a 12-person construction office where six people submit expenses each month:

The Collection Link doesn't replace the expense management lifecycle — if your team needs multi-step approval and automated reimbursement, per-user platforms still have a role. What it does is eliminate the per-user premium for the extraction layer: the part of the workflow that turns receipt images into structured data. For a small team where the founder reviews expenses personally and cuts checks manually, that extraction layer is most of what they need. The rest of the expense management suite is overhead they didn't ask for.

What Shared-Pool Pricing Actually Costs at Small-Team Volume

Shared-pool pricing means you buy processing capacity — measured in credits — rather than user seats. One credit processes one page of a document. A three-page expense report PDF with six attached receipt images consumes nine credits. A single receipt photo consumes one. The key difference from per-user pricing: credits are consumed by the work that gets done, not by the people who could do it.

At ImageToTable.ai, two shared-pool plans fit the small-team profile:

For a team processing 15 expense reports per month — a typical engineering consultancy or small agency — each report averages four to six pages (the report cover sheet plus three to five receipt attachments). That's 75 to 90 credits per month. The Growth Team plan covers that with over 2,900 credits left for other document types: invoices from vendors, bank statements for reconciliation, vendor quotes for comparison. The per-credit cost at $149 for 3,000 credits is roughly $0.05 per page — and since credits roll over month to month rather than expiring, bursty expense cycles don't translate into wasted capacity.

Contrast this with the per-user model. A 10-person team on Expensify Collect at $10 per user pays $100 per month regardless of volume — three reports or thirty, the bill doesn't move. That's $1,200 per year. On Zoho Expense Standard at $4 per user, it's $480 per year. But neither includes extraction in the sense of outputting structured data into a spreadsheet you can use outside their platform. The data stays inside their system. If you want the extracted data in your own Excel file — to import into QuickBooks, to share with your accountant, to merge with project cost tracking — you're exporting CSVs and reformatting. The extraction tool that costs $149 processes the same reports, outputs directly to Excel or CSV, and doesn't care whether three people or thirty submit — the credit draw is proportional to the work done, not the roster size.

There's a broader point here that extends beyond any single tool. For a more detailed comparison of how subscription models stack up against pay-as-you-go at different volumes, see our breakdown of pay-as-you-go vs subscription pricing across 10, 50, 200, and 1,000 pages. The pattern holds across document types: the subscription model wins only when volume is high and consistent every month. For small-team expense reports — bursty, variable, and rarely hitting the usage caps included in a subscription — a shared-pool model where credits don't expire is almost always cheaper.

The 60-Day Clock: Why Collection Speed Has Tax Consequences

IRS accountable plan rules under Publication 463 and Treasury Regulation §1.62-2 require three things for an expense reimbursement to remain tax-free: the expense must have a business connection, the employee must substantiate the expense within a reasonable period (defined as 60 days after the expense was paid or incurred), and any excess reimbursement must be returned within 120 days. If any of these conditions isn't met, the reimbursement is reclassified as taxable wages — subject to income tax withholding, Social Security, and Medicare — for both the employer and the employee.

A missing receipt doesn't just mean the employee gets a smaller reimbursement. It means the reimbursement they did receive becomes taxable income, and the employer owes payroll taxes on it retroactively. The GBTA Foundation found that companies spend an average of 3,000 hours per year correcting expense report errors — and the US Chamber of Commerce found that 40% of small businesses cited "employee time spent filling out expenses and collecting receipts" as their top expense management challenge. Every hour spent chasing a receipt is an hour the 60-day clock ticks closer to a taxable reclassification.

The IRS accountable plan rules don't prescribe a specific collection method. What they prescribe is a result: substantiation within 60 days. The faster the collection mechanism — the shorter the gap between the employee incurring the expense and the administrator receiving the documentation — the smaller the window in which a substantiation failure becomes a tax liability. A Collection Link shortens that gap because it removes the two biggest friction points in the collection process: account creation and login. An employee who incurs a $45 client lunch expense on Tuesday can upload the receipt photo on Tuesday — from their phone browser, through the link they already have bookmarked, without downloading an app or resetting a password.

This is where the cost of per-user pricing intersects with compliance risk in a way that most pricing comparisons miss. A tool that costs $100 per month in user fees but doesn't get used — because half the team never created accounts, or forgot their passwords, or finds the mobile app too cumbersome — creates compliance exposure. The per-user fee gets paid every month, but the receipts don't arrive within 60 days. The cheaper tool that's frictionless to use — the one your team actually opens — closes the collection gap. The price on the pricing page matters less than the behavior the tool produces. IRS Publication 463 doesn't care what you paid for the software. It cares whether the receipt is in the file.

When Per-User Pricing Actually Makes Sense

The argument this article makes is not that per-user pricing is always wrong. It's that per-user pricing is wrong for a specific profile: small teams where expense submission is irregular and the full expense management lifecycle is overbuilt for what you need. There are scenarios where per-user pricing is the right answer, and it's worth naming them so you can recognize whether you're in that camp or the other one.

Per-user expense management platforms earn their price when your team needs automated reimbursement — direct deposit to employee bank accounts, synced with payroll and tax withholding. If you're processing reimbursements for 30 employees every two weeks, the per-user fee buys a real reduction in administrative labor. Per-user platforms also make sense when you need multi-level approval routing — a sales director approves their team's expenses, finance reviews for policy compliance, and the controller signs off before payment. And they make sense when you're integrating with an ERP — SAP, Oracle, NetSuite — where the bidirectional data sync justifies the per-seat cost because the alternative is manual journal entries at scale.

But if you're a small team where the founder or office manager reviews expenses personally, approves with a glance, and cuts checks or initiates ACH transfers manually, the extraction layer is what you need. The rest of the suite is software you'd pay for in monthly installments and never fully use — joining the 66% of subscriptions that Spendesk found sit idle. If that describes your team, the question isn't which per-user platform to choose. It's whether you need a user platform at all, or just the extraction.

For a deeper look at how the same logic plays out across other document types in small operations, see our analyses of affordable invoice extraction for small business, receipt extraction for freelancers, and bank statement extraction for small business — the shared-pool approach applies across document types, not just expense reports.

FAQ

Can expense extraction handle reports with mixed currencies?

Yes. AI-based extraction identifies currency symbols and amounts on each line item independently. A report containing a USD hotel bill, a EUR dinner receipt, and a GBP train ticket can all be processed in a single batch. The output preserves the original currency; conversion to a base currency is a separate step you handle in your spreadsheet or accounting software. The extraction tool reads what's on the page — it doesn't perform foreign exchange.

What if an employee submits a blurry photo of a receipt?

AI extraction is more tolerant of imperfect images than template-based OCR — it reads by understanding the content semantically rather than matching pixel grids — but it has limits. A receipt photo taken in a dark restaurant with motion blur and glare will produce lower accuracy than a clear, well-lit shot. If the image is too degraded for the AI to read, you'll see missing or flagged fields in the output, and you'll need to manually verify those entries. The tool won't fabricate data to fill gaps — it will leave cells blank rather than guess.

Do I need to train the tool on my company's expense report format?

No. Unlike template-based tools that require you to define zones or train on sample documents, AI-based extraction — the approach used in ImageToTable.ai — works through what's called Custom Column Extraction: you type the field names you want (Date, Vendor, Amount, Category, Business Purpose) and the AI locates each value by understanding what it means, not where it sits on the page. This works across different report formats, different receipt layouts, and different currencies without retraining. If your team submits expenses through a standardized form, it works. If each employee uses a different format, it still works — the AI reads the contents, not the template.

Can I export the extracted data directly to QuickBooks?

ImageToTable.ai exports to Excel (XLSX) and CSV, which are both importable into QuickBooks, Xero, and most accounting platforms. There's no native QuickBooks sync — you export the file and import it through your accounting software's standard import function. This adds one manual step compared to a platform with native integration, but it also means the extraction tool doesn't need API access to your financial data — the data passes through your hands, not through a third-party connector.

Is a Collection Link secure if I'm sharing it with employees?

The Collection Link requires a verification code before anyone can upload files — you set the code, and you share it separately from the link itself. Uploaded files are processed and then deleted from the server after a processing window. The link doesn't grant access to your account, your processing history, or any other employee's submissions. Each uploader sees only the upload page. For a deeper look at the Collection Link workflow, see our guide on collecting employee expenses into Google Sheets with a Collection Link.

What's the smallest team size where extraction makes financial sense?

At five reports per month — roughly a three-person consultancy — the extraction tool costs roughly $149 per month for 3,000 credits, of which you'll use about 30. That's $149 for work that would take an hour of manual data entry. If your billable rate is above $150 per hour, it pays for itself. Below that, or if reports are fewer than five per month, manual entry might be the more economical choice. The crossover point depends on your hourly cost, but five to eight reports per month is the range where extraction starts to beat transcription on pure economics — before you factor in error reduction and the IRS 60-day compliance window.

The Bottom Line

Per-user pricing for expense management software is a convention, not a law of nature. It came from the SaaS playbook — provision seats, charge per seat — and it works when usage is proportional to headcount. For small teams processing expense reports, that proportionality breaks. Half your team submits expenses once a quarter. Two people never submit. The per-user model charges on potential, not on actual use.

The alternative is to unbundle the problem. The extraction layer — reading receipts and expense reports, identifying data fields, outputting structured results — doesn't need a per-user pricing model. It needs a per-document pricing model, because the work scales with the number of documents, not the number of people on the roster. A Collection Link makes this technically possible: one link shared with the team replaces ten individual accounts, and the processing cost tracks what gets submitted, not who could submit.

If your expense report volume is low and irregular, and your team doesn't need automated approvals or ERP integration, you're paying for software you don't use. The per-user fee you see on the pricing page is the floor — the real cost is the seats you provision for people who never open the app. Extraction that costs $149 per month and reads every report regardless of format or submitter isn't just cheaper. It matches the cost to the work, which is what pricing is supposed to do.