Home / / Product

Continuous Monitoring: Why a Fee Audit Has to Be a Subscription, Not an Event

The savings from a fee audit have a half-life. Waived charges reappear, corrected rates drift, and the next audit starts from a worse baseline than the last one finished at.

A glass office tower rising against a clear blue sky

A fee audit produces a good week. Someone finds 30,000 dollars a year, the vendor concedes most of it, the statement drops, and everybody moves on to the next thing. Eighteen months later a different person runs a different audit and finds 26,000 dollars a year.

The second number is not evidence that the first audit was wrong. It is evidence that the first audit was a snapshot of a moving system.

What decays, and how fast

Waived fees come back

A waiver is often applied as an account-level adjustment rather than a change to the underlying price. Adjustments expire, get dropped in a platform migration, or fall off when the account is re-boarded. The fee was never removed; it was suppressed, and the suppression has a shelf life.

Corrected rates drift again

The mechanics that produced the original overcharge are still running: announced increases, new fee lines, lapsed tiers, escalator clauses. Correcting a rate does not disable them. It resets the clock. The full inventory is in Vendor Fee Creep.

Your business changes underneath the pricing

Average ticket rises. Card-present mix shifts online. A new location opens on a different platform. Every one of these changes which interchange categories you land in and which fee structure suits you, and none of them triggers a review at the vendor's end unless the change costs the vendor money.

The uncomfortable arithmetic. AIME puts the cost of that drift at up to 20 percent of controllable spend. An annual audit means living at the top of that range for eleven months out of twelve.

What a monitoring cycle actually checks

Monitoring is not re-running the whole audit every month. It is a small set of comparisons against a known-good baseline, which is much cheaper and catches drift on its first appearance rather than its twelfth.

  • Line-for-line against last period. Any charge that changed rate, any charge that appeared, any charge that disappeared. New lines are the highest-signal event on a statement.
  • Against the negotiated position. The rates agreed after the last audit become the baseline, so a reversion is caught as a reversion rather than rediscovered as a finding.
  • Effective rate trend. A rising effective rate on flat volume and flat mix is drift by definition. If the mix did move, the trend should be explainable by the mix, and if it is not, something else moved.
  • Pass-through against published tables. Network rates change on a schedule. Yours should change on the same schedule, by the same amount, and not otherwise.

Why nobody does this manually

Because it is genuinely dull and genuinely skilled at the same time, which is the worst combination for getting a task done inside a busy finance team. It needs someone who can read a statement, remember what last month said, and care about a three dollar line. It has to happen every month, for every vendor, forever, and it produces nothing at all in the months where nothing changed.

That is exactly the shape of work to hand to an agent. The months where nothing changed cost nothing to check. The month something moves is the month the whole exercise pays for itself.

What good looks like

A monitoring loop should be judged on how quickly it catches a change, not on how much it finds. A program that finds 40,000 dollars a year every year is not succeeding; it is failing on a twelve-month cycle. A program working properly finds a large number once and then small numbers forever, because it is intercepting drift at the point it starts.

Frequently asked questions

Does monitoring mean switching vendors constantly?

The opposite. Most drift is corrected by the incumbent when it is documented, because it is easier for them to reprice than to lose the account. Monitoring makes staying viable, since the reason businesses switch is usually that the price got away from them unnoticed.

How many months of history do I need to start?

One statement gives you a benchmark position. Two or three consecutive months of the same vendor give you a trend, which is what turns a benchmark into evidence of drift.

Does this apply outside card processing?

Yes, and the mechanics are nearly identical. Bank treasury fees, freight and parcel invoices, and software renewals all drift for the same structural reasons. See the bank fee audit guide, the freight audit guide and the SaaS spend audit.

What does it cost to watch a vendor?

Running the audit is free, and AIME is paid out of what it actually recovers. That alignment is deliberate: a monitoring program billed as a retainer gets paid the same in the months it finds nothing.

See what your statements are hiding.

Drop a statement into AIME and get a line-item savings analysis in minutes. No retainer, no hourly fees: you pay only when AIME recovers savings.

Run a free live audit

Keep reading

Vendor Fee Creep: Why Your Contracted Rates Quietly Rise Every YearFee creep is the slow drift between the price you negotiated and the price you pay. How it works across processing, banking, freight and software, and how to stop it. Contract vs. Statement: The Comparison Nobody Has Time to DoYour merchant agreement says one thing and your statement charges another. How to run a contract-to-invoice comparison, what drift looks like in practice, and why rounding is where money hides.