What a 5% bad-record rate actually costs at five million sends

Ask a demand-gen lead what bad data costs them and you will almost always get the purchase price back: so many records, at so much each, of which some percentage turned out to be junk. It is an honest answer, it is easy to defend in a budget meeting, and in our experience it accounts for a small fraction of the real number.

The rest of the bill is real. It just never arrives in a format anyone files.

The three bills

The first bill is the data. You paid for rows that were never usable. This one is visible, it sits on an invoice, and it is genuinely recoverable — but only if you can prove which rows were bad and why. That proof is an audit trail, and most teams asking for a credit note do not have one, which is why most of those conversations go nowhere.

The second bill is your team's time. Every dead row that reaches an SDR gets the full treatment before anyone gives up on it: research, a personalised first line, a sequence slot, two or three follow-ups, a note in the CRM. At volume this is usually the largest single cost in the whole exercise, and it is invisible for a boring reason — it is spread across salaries you were paying anyway, so it never shows up as a line item. Nobody ever gets an invoice for the eleven hours an SDR spent on people who left last year.

The third bill is sender reputation, and it is the one that compounds. Mailbox providers weigh engagement as heavily as bounces. A list of technically valid but humanly dead addresses produces deliveries with no opens, no replies and no positive signal (and if you are measuring those numbers at all, measure them net of proxies and scanners) — and that pattern, repeated across campaigns on the same domain, moves you steadily towards the spam folder for the recipients who would have engaged.

Why the third bill is the dangerous one

The first two bills are linear. Twice the bad data, twice the cost, and you can reason about them on a spreadsheet. The third does not behave like that. Sender reputation degrades gradually and recovers slowly, so by the time it is visible in your numbers the damage was done several weeks earlier, in campaigns you have already stopped thinking about. Recovery is measured in months of throttled volume and warm-up, not in a credit note.

This is the part clients tend to accept intellectually and underweight in practice, because it is the only one of the three with no date on it.

Deliverability is not a setting you configure. It is an outcome of what you have been sending, to whom, for the last ninety days.

Where the leverage is

Because the third bill compounds, the value of catching a bad row collapses the moment it is sent. That ordering — not the per-record price of validation — is the whole argument for doing this upstream of the send:

• Before the send — you pay a per-record validation cost and nothing else.
• After the send — you have already spent the SDR time and already taken the reputation hit.
• After the quarter — you are rebuilding reputation, which costs sending volume you cannot get back.

Put a number on your own list

We are not going to give you an industry average here, because the honest answer to “what is this costing us” is that it depends entirely on your list, your sources and your ICP, and no vendor's benchmark — including ours — will tell you anything useful about your own file.

What does tell you is running a real campaign file through validation and reading the reasons column: how many contacts have left, how many were never employable targets, how many were sitting on your own suppression list already. We have watched that number surprise people in both directions.

Either way it is a measurement rather than an estimate, and it takes about a day.