Published benchmarks for cost per lead are close to meaningless, because they average across deal sizes that differ by two orders of magnitude and definitions of "lead" that differ by more. The number you need comes from your own deal value, win rate and margin, and it takes about ten minutes to work out.
Work backwards from a deal
Take your average contract value and the margin on it. Decide what proportion of that margin you are willing to spend to acquire one. Then divide by the rate at which qualified leads become customers.
If a qualified lead converts at one in five and you will spend a fifth of the first year margin to win one, the ceiling on a qualified lead is that margin divided by twenty-five. That is your number. It is not anybody else’s.
A worked example
Take a company with a 40,000 average first year contract and a 60 per cent gross margin, so 24,000 of margin per customer. They are willing to spend a quarter of that to win one, which is 6,000 of acquisition budget per closed deal.
Qualified leads convert at one in five, so five leads are needed per customer. Six thousand divided by five puts the ceiling at 1,200 per qualified lead. If outbound is producing them at 700 the channel has room to scale; at 1,800 it is losing money on every one and the problem is either the conversion rate or the definition of qualified, not the campaign.
Those figures are illustrative. Substitute your own contract value, margin and win rate and the arithmetic is the same: margin, times the share you will spend, divided by the number of qualified leads it takes to close one.
Two definitions have to be stable
The figure is only comparable if "qualified" means the same thing this quarter as last, and if the same person applies it. Most apparent swings in cost per lead are definition drift rather than performance change.
Agree the grade, write it down, and keep it fixed for at least two quarters. A metric that moves because its denominator moved is not telling you anything about the campaign.
Compare across channels, not against the industry
The useful comparison is between your own channels over the same period: outbound against events against paid against partnerships. That tells you where the next pound goes.
A channel that produces expensive leads may still be worth running if those leads close faster or at higher value, which is why cost per qualified lead should never be read without the win rate beside it.
When the number is high and that is fine
Long cycle, high value sales produce expensive leads and should. A six figure contract with an eighteen month cycle can justify a cost per lead that would be absurd for a self-serve product.
The failure mode is not a high number. It is a number nobody has calculated, which means every channel discussion is conducted on instinct and the loudest channel wins.