Client refunds
How to reduce client refunds without tightening the contract.
A refund is earned revenue leaving quietly, months after the thing that caused it. Tighter terms win the argument and lose the client. The durable fix sits upstream, in delivery.
Why clients actually ask for refunds
Almost never because the work was bad. In the refunds we've traced back through client delivery pipelines, the trigger is expectation drift: the client expected something the contract didn't say, nobody caught it early, and the distance between the two grew until it became a conversation about money.
The important part is the lag. The delivery failure that causes a refund is usually two to three months older than the request. By the time it reaches you it looks like a billing dispute, which is why most agencies respond by rewriting their terms and see no change in the rate.
Most B2B agencies sit between 3% and 8% of revenue once credits, partial refunds, and goodwill discounts are counted. Ask an owner and they'll usually name half that, because the quiet discounts never got filed as refunds.
The causes
Four failures that turn into refund requests
Expectation drift after kickoff
The client expected something the contract didn't say, and nobody corrected it in week one. Every week after that the gap widens, because both sides are now working from a different picture of what success looks like. This is the single most common refund cause we find.
Silence at the moment of doubt
Clients rarely ask for a refund while they're being talked to. They ask after a stretch of not hearing anything, when the invoice arrives and they can't recall what the last month bought them. The delivery may have been fine. The perception of it was not.
Handoffs that drop context
Work moves between strategist, specialist, and account lead, and something agreed early doesn't survive the trip. The client notices, because they're the only person present at every stage. From their side it reads as an agency that isn't listening.
Results arriving later than implied
Not overselling exactly, but a timeline the sales conversation left comfortably vague. The client anchors on the optimistic read, delivery lands on the realistic one, and month three is where that difference gets invoiced back to you.
What to fix, in order
Measure the real rate first. Twelve months of refunds, credits, discounts, and write-offs as a percentage of revenue, then split by segment and by service line. If they cluster in one segment, the issue is who's being sold to. If they're spread evenly, it's delivery.
Then fix onboarding, because it sets the expectation every later month is judged against. Say plainly what the client gets, when, and what it will not include. The conversation that feels awkward in week one is the refund you don't have in month four.
Then instrument the quiet stretches. Find the points where clients historically go silent and put a real check-in there before the invoice does it for you. Most refund requests are preceded by three or four weeks of nothing.
Fixing the refund leak usually returns more margin than any price rise the same agency was considering. The wider picture is on our agency profitability page, and the numbers from past engagements are in our case studies.
What it costs
Half of what we save you, once.
Thirty minutes on your refund rate and your delivery pipeline gives us the projection: what the leak costs per year and what closing it is worth. Our fee is 50% of that number, paid once. You keep the projection whether or not we work together.
Questions agency owners ask about refunds
Rarely because the work was bad. In most agency refunds the trigger is expectation drift: the client expected something the contract didn't say and nobody corrected it early. By the time the refund request arrives, the delivery failure that caused it is usually two or three months old.
Most B2B agencies sit between 3% and 8% of revenue once write-offs, partial credits, and quiet discounts are counted. Owners routinely estimate their own rate at half the real number, because partial credits and goodwill discounts never get recorded as refunds.
Tighter contracts reduce your legal exposure and do nothing for the relationship, so you win the argument and lose the client and the referral. The durable fix is earlier detection: set expectations explicitly at onboarding, put checkpoints where the work usually goes sideways, and reach out at the moments clients typically go quiet.
Both, and the split matters. If refunds cluster in a particular segment, sales is selling to people the delivery process can't serve. If they're spread evenly, it's delivery. Segmenting refunds by source before changing anything is what tells you which one you have.
More than the refunded amount. Every refund carries the delivery cost you already paid, the referral you won't get, and the team hours spent handling it. A 5% refund rate on $3M in revenue is rarely a $150K problem once those are counted.
We record the baseline rate before anything changes, then track it monthly against that number, split by the cause we identified. Refunds lag their cause, so the honest read comes a quarter out rather than in the first few weeks.
Ready to see the math
Your bottom line has room. We can show you where.
Book a free 30-minute call. We'll look at your refund rate and growth trajectory, then show you the savings we'd project. No pitch deck, no commitment.
Free 30-minute call. The math is yours to keep either way.