Travel Invoice Reconciliation Software: How TMCs Close Faster

Ask a TMC finance manager what happens at month end and you will hear a version of the same story: a spreadsheet, three data sources that do not agree, and several days spent working out why.

Travel reconciliation is unusually painful compared to other industries, for structural reasons that are worth naming precisely — because the reasons determine which software actually helps.

Why travel reconciliation is harder than normal AP

In most businesses, an invoice matches a purchase order matches a delivery. In travel, a single trip can generate:

• A GDS booking record with one fare
• A supplier invoice with a different total, because taxes and surcharges were calculated later
• A card transaction that may be split, aggregated with other bookings, or posted weeks after travel
• One or more change or cancellation transactions, each with its own fee
• A commission or override due back from the supplier
• A client service fee applied per the TMC’s own contract

Six artefacts, three systems, and multiple legitimate reasons why the numbers differ. Multiply that across thousands of monthly transactions, and reconciliation stops being a clerical task and becomes forensic work.

Then there are the genuinely hard cases:

Partial refunds and residual value. A traveller changes a ticket. Part of the original fare carries forward, part becomes a residual credit, and a change fee applies. Matching the resulting transaction back to the original booking defeats most rules-based systems.

Timing mismatches. The booking was made in one period, travel occurred in another, and the supplier invoiced in a third. Which period does it belong to?

Aggregated card postings. A single card settlement line covering fourteen bookings, which must be split back out before anything reconciles.

What reconciliation software should actually do

The category label covers a wide range of capability. These are the functions that matter for a TMC specifically.

Automated three-way matching

The core function: matching booking record to supplier invoice to payment, automatically, with tolerance thresholds. Small variances within a defined band clear without human review; anything outside it is flagged.

Getting tolerance right is important. Set it too tight and finance drowns in trivial exceptions. Too loose and real errors clear silently. Good systems let you set tolerance per supplier — because a small variance from one airline is rounding, and from another it is a systematic error worth pursuing.

Exception queues that prioritise by value

When something does not match, it should arrive in a queue ranked by financial impact, not chronologically. Finance teams have finite time; they should spend it on the largest discrepancy before the trivial one.

Supplier commission and override tracking

Money owed to the TMC is where reconciliation most often leaks. Commissions that were never paid, overrides not applied at the agreed tier, incentive thresholds missed by a small margin. This revenue is invisible unless something is tracking it deliberately, and it rarely appears in a general-purpose AP tool.

Automated period close

The output of reconciliation is a close. Software should produce the journal entries, accruals for travel booked but not yet invoiced, and the reporting pack — not just a list of matched transactions for someone to re-key.

A complete audit trail

Every match, exception, override and manual adjustment logged with who did it and why. This matters for client audits and for the TMC’s own year-end.

Rules-based matching versus learning systems

Traditional reconciliation tools use static rules: match on record locator, then passenger name, then amount within tolerance. These work well for clean transactions and fail on exactly the cases that consume the most time — the partial refunds, the aggregated postings, the supplier who changed their invoice reference format.

Systems that learn from how your team resolves exceptions handle this differently. When a finance controller manually matches an awkward residual-value transaction, that resolution becomes a pattern the system applies next time. Match rates improve over time instead of remaining flat.

This is the practical difference between reconciliation software that plateaus at 70% auto-match and one that keeps climbing. TMCs running the VantisCorp Finance Agent reconcile roughly 3x faster than their previous manual process, with the finance team’s time redirected from matching to investigating genuine exceptions.

As Sarah Mitchell, COO of Premier Travel Group, put it: their finance team “closes books in hours instead of days.”

What to check before you buy

Does it read your actual data sources? Not “supports GDS integration” in the abstract — your GDS, your mid-office, your card provider’s file format. Ask for confirmation against your specific stack.

How does it handle changes and refunds? Ask the vendor to walk through a partial refund with residual value and a change fee. This single scenario separates capable products from demos.

What is the realistic auto-match rate? Push past the marketing number. Ask what it is at go-live versus after six months, and on what transaction mix.

Who owns unmatched items? Some systems flag exceptions and stop. You want workflow: assignment, ageing, escalation.

Can it handle multi-entity and multi-currency? If you operate across markets, currency gain/loss handling on reconciliation is not optional.

A realistic implementation sequence

Reconciliation is a good automation candidate because success is unambiguous — either the transactions match or they do not.

• Baseline first. Measure current auto-match rate, days to close, and hours spent. Without this you cannot demonstrate improvement, and you will be arguing about the value of the project six months in.
• Start with one supplier or one client. Prove the matching logic on a contained data set.
• Run parallel for one cycle. Reconcile manually and automatically for a single month and compare. This catches encoding errors before they affect a real close.
• Expand by volume. Add your highest-transaction suppliers next, since that is where hours concentrate.
• Then automate the close itself. Only once matching is trusted should you let the system generate journal entries.

Expect the first cycle to be slower, not faster. Teams that abandon reconciliation automation usually do so in week three, before the learning curve pays back.

The wider point

Reconciliation is often treated as a back-office cost to be minimised. For a TMC it is closer to a revenue function: it is where unpaid commissions get caught, where supplier billing errors get identified, and where the margin on each transaction is actually confirmed rather than assumed.

Automating it does not just save finance-team hours. It makes the TMC’s own economics visible in something closer to real time — which is a prerequisite for pricing new business accurately.

---

See it in practice: the Finance Agent automates three-way matching, tracks commissions and overrides, and produces period close in hours rather than days — connected to the GDS, mid-office and finance systems you already use. Review our integrations.