Expense approval process flowchart

What the expense approval process is

An expense approval process deals with money an employee has already spent. Someone paid for a train ticket, a client dinner or a hotel, out of their own pocket or on a company card, and the organisation now has to check it, code it and settle it. That makes it a different problem from invoice approval, which handles a supplier's claim against a purchase order. Here there is no order to match against and no supplier record to check; the evidence is a receipt only the claimant holds, the counterparty is a colleague who is out of pocket until finance acts, and the money usually leaves through payroll rather than the accounts payable payment run.

Claims stall for a short list of reasons. The receipt is missing or unreadable. The claim is coded to a category or cost centre that does not carry the spend. The claim is out of policy, and gets treated as a data error and bounced back for correction when it is really a decision somebody with authority has to make and record. Approval is given by whoever is nearest rather than by whoever holds the budget. And nobody looks for duplicates, which in expenses means something quite different from invoices: a taxi claimed by both people who shared it, a card transaction claimed again as cash, a receipt photographed twice and submitted a month apart.

This template maps a claim across five lanes, Employee, Line manager, Finance / AP, Payroll and Auditor, and seven phases from capture to assurance. It is deliberately explicit about the three points teams argue over afterwards: the policy check has separate exits for a correctable error and a genuine breach, approval routing is driven by a stated value threshold rather than by who is available, and the duplicate check runs before any coding or payment effort is spent. Change the thresholds, categories and role names to match your own expenses policy and delegation of authority.

What this flowchart covers

In this template

• Five swimlanes, Employee, Line manager, Finance / AP, Payroll and Auditor, across seven phases: Capture, Submission, Validation, Approval, Finance review, Reimbursement and Assurance.
• A single "Receipt attached and within policy?" gate with three exits: a compliant claim goes to the line manager, an incomplete one is returned to the claimant to resubmit, and a policy breach is sent for an exception decision instead of being bounced back as an error.
• The breach branch at "Exceptional approval granted?" in the Finance / AP lane. A granted exception rejoins the normal approval route; a refusal ends at "Claim rejected and closed", the same exit a line manager rejection uses.
• Two approval gates in sequence: "Line manager approves claim?", then "Value over approval threshold?", where a claim above the limit picks up "Escalate to second approver" before finance sees it.
• A "Duplicate claim detected?" check as the first act of finance review, with a "Hold and query duplicate claim" exit, followed by "Check evidence and code VAT" and "Post claim to expense ledger".
• Two reimbursement routes at "Reimburse via payroll or payment run?" that converge on "Employee reimbursed and notified", plus a sampled branch from posting into the Auditor lane: select claims for sample audit, test whether each is supported by evidence, recover and log anything that is not, and report the findings.

When to use this template

• Rewriting an expenses policy, where the routing and the exception path are easier to settle as a picture than as paragraphs.
• Configuring an expense management tool or ERP module, where approval routing, thresholds and receipt rules have to be specified before they are built.
• Setting or reviewing the value at which a second approver is required, so the delegation of authority is a visible routing rule rather than habit.
• Briefing new managers who approve claims, and new claimants who need to see what happens when a receipt is missing or the spend is out of policy.
• Preparing for an expenses review or a VAT check, where you need to show how claims are evidenced, coded and sampled after payment.

How it works

1. Rename the lanes to the roles you have

   Replace Employee, Line manager, Finance / AP, Payroll and Auditor with your real functions. In smaller organisations finance and payroll are one desk: merge those lanes rather than drawing a hand-off that never happens. If a shared service centre pre-checks claims before the manager sees them, give it its own lane instead of leaving that work in the manager's.

2. Write down what "within policy" means

   The check is only as good as the policy behind it. List the claimable categories, the per-head limits, the value below which a receipt is not required, and the mileage and per diem rates. Attach that list to the "Receipt attached and within policy?" step so every approver applies the same test.

3. Put real figures on the threshold

   Replace "Value over approval threshold?" with your own limit, say which currency it is in, and state who signs at each band. Decide whether the test reads the claim total or a single line, or one trip will arrive as several claims that each sit under the limit.

4. Define the duplicate check keys

   Decide what makes two claims the same. Claimant, date, amount and merchant together is the usual combination, widened to compare across claimants for shared costs such as a taxi or a team meal. State that a suspected duplicate is held and queried rather than deleted, because genuine repeat spend on the same route on the same day does occur.

5. Choose the reimbursement route and publish the cut-off

   Decide which claims go through payroll and which go in a payment run, and publish the cut-off for each. Most expense disputes are about timing rather than amount: a claimant who knows the weekly cut-off and the next payroll date stops chasing finance.

6. Agree the sampling rule, then keep one current version

   State how often the audit sample is drawn and how it is weighted, for example towards high-value claims, granted exceptions and approvals given by someone who reports to the claimant. Walk the finished chart through with a manager, finance and payroll, correct it to what they actually do, then publish that revision and keep the earlier ones, so anyone opening the chart later can tell which version they are reading.

Frequently asked questions

What are the steps in an expense approval process?

A typical sequence is: the employee incurs the expense and captures the receipt, then submits a claim coded to a policy category and a cost centre. The claim is checked for a valid receipt and for policy compliance, with incomplete claims returned to the claimant and out-of-policy claims sent for an exception decision. The line manager approves the business case, a value threshold decides whether a second approver is needed, and finance then checks for duplicates, verifies the evidence, codes the VAT and posts the claim. Reimbursement follows either in the next payroll cycle or in a payment run, and a sample of approved claims is audited afterwards.

How is expense approval different from invoice approval?

An expense claim is employee-incurred spend reimbursed to a person; an invoice is a supplier's demand for payment, usually raised against a purchase order. The practical differences follow from that. There is no purchase order or goods receipt note to match against, so the evidence is the receipt plus the business purpose the claimant states. There are no payment terms, so the deadline is the payroll or payment-run cut-off rather than an invoice due date. The approver is judging whether the spend was reasonable and within policy, not whether an amount agrees to an order. And duplicate detection has to work without an invoice number, on claimant, date, amount and merchant instead.

Who should approve an expense claim, and how do value thresholds work?

The claimant's line manager should approve every claim, since they are the one placed to judge whether the trip or the purchase was necessary and whether the stated business purpose holds up. Nobody approves their own expenses, and a claim should not be signed off by someone who reports to the claimant. Above a stated amount a second approver joins, typically a senior manager or the finance director, so larger claims get a second reading without slowing routine ones down. That limit belongs in the expenses policy with the routing set to enforce it, so no approver is weighing up on the day whether a claim is big enough to pass on.

What should happen to an out-of-policy expense claim?

Treat it as a decision rather than an error. Returning it for correction invites the claimant to recode the same spend into a category that passes, and it leaves no record that the breach happened. In this chart the breach has its own exit into "Exceptional approval granted?": someone with the authority to make an exception either grants it, with the reason recorded, in which case the claim rejoins the normal approval route, or refuses it and the claim is rejected and closed. Review granted exceptions periodically, because a category that needs an exception every month is a policy that needs changing.

How do you prevent duplicate expense claims?

Run the check as the first step of finance review, before any coding or payment work is spent on the claim. Expense claims have no invoice number, so match on the combination of claimant, date, amount and merchant, and compare across claimants as well as within a single claim: a shared taxi or a team meal is often submitted by two people who both attended. Watch for card transactions claimed again as cash and for receipts re-photographed and submitted weeks later. Hold and query a suspected duplicate rather than deleting it, because genuine repeat spend, such as the same journey on the same day each week, does happen.