The hidden math behind ROI in B2B appointment setting

"Our campaigns average 8–12× ROI." You have read that line on every agency site you have shortlisted, including ours in spirit. It is close to meaningless on its own, and the reason is not that the agencies are lying. It is that a return multiple is the output of four assumptions, and nobody shows you the four.

Here they are.

The four inputs, and who owns each

Any appointment-setting ROI number is this:

(meetings × show-up rate × your close rate × your deal value) ÷ what you paid

• Meetings: the agency's number.
• Show-up rate: shared. The agency's confirmation process moves it; so does how attractive your meeting is to attend.
• Close rate: yours. Entirely.
• Deal value: yours.

Two of the four belong to you, and they are the two with the widest possible range. Which is why a vendor's headline multiple tells you about their best client's funnel, not about yours.

Where the number usually goes wrong

Booked versus held. A forecast built on booked meetings and a reality made of held ones diverge by whatever the no-show rate is. Ask which unit the number uses. Ask what happens to a no-show contractually. That answer tells you more about a vendor than the multiple does.

Qualified versus interested. A meeting with someone who cannot buy is not a smaller version of a meeting with someone who can; it is a different thing wearing the same label. Get the qualification standard in writing before the engagement, not after the first disputed invoice.

The close rate borrowed from inbound. This is the most common error we see in a buyer's own spreadsheet, and it is theirs, not the vendor's. Outbound meetings close at a lower rate than inbound ones, because the buyer did not raise their hand. Using your inbound close rate to model outbound will overstate the return by a wide margin, and then the campaign gets blamed for the spreadsheet.

First-year revenue against multi-year value. If your contracts renew, first-year revenue understates the return. If they churn at 40%, lifetime value overstates it. Pick one and say which.

The ramp. Almost every published multiple is a steady-state number. The first weeks are list-building and message testing, and the meetings land behind that. A twelve-month projection that starts at full rate in week one is wrong by roughly a month of pipeline.

How to check a vendor's number in five minutes

Take their claimed multiple and work backwards:

1. Ask what deal value it assumed. If it is materially above yours, the multiple does not transfer.
2. Ask what close rate it assumed, and whether that was outbound-sourced.
3. Ask whether the meeting count is booked or held.
4. Substitute your own numbers into the same formula.
5. Compare the result to what you are being quoted.

If a vendor cannot answer the first three, the multiple was decoration. If they can, you now have a model you can actually defend to a CFO, which is the real point of the exercise.

What we would rather you do

Model it yourself, with your figures, before you talk to anyone, us included. The ROI calculator takes deal value, close rate and target meeting count and returns the arithmetic above with nothing hidden in it. It defaults to the conservative end on purpose.

Then bring the output to the pricing conversation. A quote argued against your own model is a much shorter conversation than a quote argued against a number from someone's homepage.