Corporate Expense Management White Paper: The 2026 Guide

Before discussing strategy, let us size the problem. According to a GBTA Foundation study conducted with HRS, processing an expense report costs an average of $58 (approximately €53) and takes 20 minutes to complete^[1]. And that is just the baseline: nearly one in five reports contains an error, the correction of which costs $52 and an additional 18 minutes^[2]. For a 200-employee company where each employee submits one to two reports per month, the administrative cost quickly reaches six figures per year, for a task that adds no value for the customer.

Fraud makes the bill worse. The French Institute for Fraud Prevention (Institut français de prévention à la fraude) estimates that 89% of internal fraud occurs through expense reports, at a cost of approximately €700 per year per employee^[3]. Only one in three employees would never commit expense fraud^[4], and 22% of French employees reportedly falsify their reports on a regular basis^[5]. Yet adoption remains minimal: barely 3% of SMEs use a corporate card issuance solution^[6]. This gap explains the return on investment of a properly managed expense programme.

1. Expense reports. The original silo, and often the most costly. A manual expense report chains together data entry, paper receipt collection, verification, managerial approval, accounting integration, and reimbursement, each step adding delays and the risk of error. Digitisation reduces the unit cost from €53 to €5–10^[1], a five- to tenfold improvement. In 2026, the question is no longer whether to digitise, but how to connect expense reports to the rest of spending so they never become a silo again. Full details in our guide to choosing an expense management solution with a payment card.

2. The corporate card. This is the lever that eliminates the expense report at source. A corporate or virtual card, assigned to the right employee with a spending limit and category, transforms a purchase into a structured transaction at the point of payment: amount, date, merchant, category. It is all there, with nothing to enter manually. The classic trap is opening a separate card for each use case (travel, online purchases, fuel) and recreating the very fragmentation you set out to avoid. A single card covering all use cases, governed by a configured spending policy, is far superior to a collection of siloed cards. For a comparison, see our selection of leading smart corporate payment cards.

3. Centralisation. Without it, the first two pillars remain islands. Centralising means bringing corporate cards, expense reports, advances, and reimbursements into a single unified system where every transaction shares the same data structure: employee, cost centre, accounting category, project, analysis dimension. This uniformity is what makes spending manageable. As long as data lives across three separate tools, no reconciliation is reliable. Once it converges, finance teams regain the visibility they previously had to reconstruct manually with pivot tables.

4. Accounting automation. This is the link that turns centralisation into tangible savings. Optical character recognition (OCR) extracts the amount, date, VAT, as well as the VAT number from a photographed receipt. Categorisation rules automatically assign the correct accounting line. Native export to the ERP or accounting software (Sage, Cegid, Pennylane, among others) eliminates manual re-entry. On this point, integration quality comes down to one simple criterion: a validated transaction must appear, without any human intervention, as the correct accounting entry in the software. If it requires a CSV export and a manual import, the automation is only partial.

5. CSR and scope 3. The most recent pillar, and the most underestimated. The CSRD now requires reporting of value chain emissions^[7], and two GHG Protocol categories directly concern corporate spending: category 1 (purchased goods and services) and category 6 (business travel)^[8]. These emissions are calculated from purchasing and mobility data, exactly the data that centralisation has already structured. A categorised transaction feeds a carbon footprint. A fragmented one never can. This is Greenway's angle on the corporate card in the service of CSR: sustainability starts with actionable spending data.

Each pillar, taken in isolation, has its own market, its own cards, and its own back-office. That is precisely the problem. A typical SME stacks: a bank card for executives, an expense report tool for employees, a spreadsheet for advances, another for budgets, and a manual bank reconciliation at month-end: five interfaces, five exports, five sources of error. Fragmentation is not an organisational detail: it is a direct cost (€53 per expense report) and a data loss (you cannot manage what you cannot consolidate).

This is the thesis of this white paper, and Greenway's approach: a single corporate card covering all spending use cases, connected to a management back-office, with native accounting integration and native impact reporting. Every transaction feeds the same unified system, and 1% of each transaction is donated to the Greenway Foundation through the 1%ForAll® programme. Spending data becomes both manageable and measurable, something no stack of spreadsheets can achieve.

A spending policy is the document that defines what is reimbursable, within what limits, and according to what process. Too often, it exists on paper but not in the tool: the finance controller checks compliance manually, line by line. The result is predictable: inconsistent review, growing delays, and fraud that flourishes in the blind spots. With 22% of French employees admitting they regularly falsify their expense reports^[5], fraud is not a marginal case but a structural issue that only structured data can contain.

Automation changes the picture on two levels. First, it codifies the policy: a breached spending limit, a rejected category, or a detected duplicate triggers an automatic block or a request for justification. Second, it makes fraud detectable: a receipt submitted twice, an amount systematically rounded just above the approval threshold, an unknown supplier. These are all signals that manual reconciliation misses but that a unified system surfaces immediately. Corporate expense management does not eliminate fraud through moralising. It reduces it by eliminating blind spots.

Many solutions market themselves as "digitised" when they merely replace a spreadsheet with a web interface, with a CSV export at the end of the chain. The real criterion is native accounting integration: a validated transaction must generate, without any manual step, the correct accounting entry in the company's software. This is where the difference lies between an expense report tool and a truly integrated expense management solution.

In practice, three things need checking. Connectivity: does the solution offer a native connector to Sage, Cegid, or Pennylane, or does it rely on a flat file that must be imported manually? Mapping depth: does each category map to the correct general ledger account and counterparty, with the right VAT rate, or is the export a uniform block that accounting must re-split? Traceability: once an entry has been posted, can you drill back to the original receipt? On these three points, the gap between a superficial and a deep integration represents days of accounting work per year.

OCR is the component that makes the whole thing credible: photograph a receipt and obtain (without any manual entry) the amount, date, VAT rate, and supplier tax number. That is what eliminates the 20 minutes of the manual expense report^[1]. But OCR only has value when connected to accounting mapping: extracting data only to re-enter it elsewhere serves no purpose. It is the complete chain (capture, categorisation, integration) that delivers the gain, not any single component in isolation.

The total cost of ownership of an expense management solution cannot be read from the licence price alone. It includes the cost of non-digitised expense reports, the accounting time spent on reconciliation, undetected fraud, and the cost of manually reconstructing CSRD reporting. Conversely, the return on investment is not limited to time savings: it includes discounts obtained through centralised purchasing, fraud prevented by automated controls, and the value of spending data that is ready for carbon footprint reporting.

On time alone, the calculation is stark. At €53 and 20 minutes per manual expense report^[1], a company processing 1,000 reports per year spends over €50,000 and 330 hours on a task that adds no value. Digitisation, which reduces the unit cost to €5–10, frees up most of that budget. The payback threshold for a paid solution is reached after the first few hundred reports, and that is without counting fraud prevented, which adds several hundred euros per employee per year^[3].

The trap, once again, is local evaluation. A finance director who only measures the licence cost misses half the ROI. The right question is: what does fragmentation cost today (in time, fraud, and reporting overhead) and how much of that does centralisation recover? From this perspective, centralising spending flows and payment methods in a single unified system is not an expense but an investment whose return is measurable in months.

To move from theory to action, seven concrete checks are enough to avoid the most costly pitfalls:

Expense management operates within an increasingly demanding regulatory environment. Mandatory B2B electronic invoicing and the CSRD transposed into French law make spending data both legally required and structured^[7]. On the tax side, the requirement for documented recoverable VAT drives businesses to retain, for every deductible expense, a readable, dated, and categorised receipt, exactly what OCR coupled with a centralised system produces.

But it is the CSRD that changes the picture most dramatically. Value chain emissions reporting covers, under the ESRS standards, purchased goods and services (GHG Protocol category 1) and business travel (category 6)^[8]. Both categories are calculated from spending data itself: a supplier purchase, a train ticket, a hotel stay. Centralised spending feeds directly into carbon reporting. Fragmented spending must be reconstructed each year from spreadsheets. Compliance becomes a competitive advantage when the data already exists in the back-office.

The same principle applies to mobility: a multi-network fuel card that is properly categorised feeds both accounting and scope 3 reporting without double entry. Regulation simply reveals a management truth: well-structured spending serves accounting, financial control, and non-financial reporting alike, provided the right unified system has been put in place from the outset.