How much does a B2B lead cost and what drives the price

Why industry averages are useless here

Reports scatter by an order of magnitude, and the cause is not research quality. Under the word lead people mean completely different objects: a click on a form, a downloaded file, a reply in a chat, a qualified opportunity. Until the definition is fixed, there is nothing to compare.

The second cause is cycle length. In markets where a decision takes months, part of this month enquiries turns into revenue half a year later, and dividing current spend by current enquiries produces a deliberately inflated figure.

The third is demand structure. In a narrow niche there are physically few potential buyers, so contact cost is high while deal economics are excellent. In a broad one it is the reverse. That is why the only meaningful average is your own, measured over several months.

What belongs in the numerator

Everything, not only the advertising budget. Agency fees, tool subscriptions, and the working time of the person who processes enquiries all belong in the cost. Leave the last one out and a channel where a human does all the work looks free.

A useful discipline is to price that time at the rate you actually pay. If a salesperson spends a third of the week sorting enquiries, a third of that salary sits in your lead cost whether you count it or not.

One thing that does not belong: the cost of building the product or the website. Those are not acquisition costs for a specific channel, and folding them in makes every channel look equally bad.

What belongs in the denominator

Only enquiries that reached a first substantive contact. A form fill with a fake phone number is not a lead, however much you paid for it. Track the rejected share as a separate number: that is usually where the gap between a pretty report and reality hides.

Be explicit about the stage you are counting. Cost per enquiry, cost per conversation and cost per qualified opportunity are three different numbers, and quoting one while thinking about another is the most common mistake in this calculation.

Count monthly and over at least a quarter. A single month proves nothing: seasonality, one salesperson holiday and one large deal distort the picture more than any channel does.

How chat monitoring is priced

The mechanism differs from advertising. You do not pay per contact; you pay a subscription for a number of sources under watch, and the cost per request is simply that subscription divided by the requests found in the month.

Two consequences follow. In a broad direction with dense demand the cost per request falls quickly, because the numerator is fixed while the denominator grows. In a narrow one it stays high, and no amount of tuning changes that: the market does not contain the requests.

The honest version of this calculation therefore always names the denominator. A price per request quoted without saying how many requests it assumes is a marketing number, not an economic one.

What to do when your figure looks too high

The first reaction is usually wrong: cutting the budget. That reduces volume, not price, because the cause almost always lies after the enquiry rather than in the channel. Check in reverse order: handling first, then filtering, then the channel itself.

Second, split enquiries by type. A high average often turns out to be two groups: cheap ones that never buy and expensive ones that do. Then the task is not to lower the average but to stop paying for the first group.

Third, be clear about which figure the period distorts. Cost per lead itself does not involve revenue, so a short window does not inflate it; what a short window distorts is cost per deal and payback, because the spend has landed and part of the revenue has not. Mixing the two is how a healthy channel gets shut down.

A worked example

Suppose you need 5 deals a month. Of the conversations you hold, 25 out of 100 become deals. Of the enquiries you answer, 20 out of 100 become conversations. Divide, do not multiply: one deal requires 20 enquiries, so 5 deals require 100.

Now take the spend. If everything included costs the same as 2 salespeople days a week plus the subscription, divide that total by 100 and you have your cost per enquiry. Compare it with the profit on one deal, not with a figure from an industry report.

The rule of thumb worth remembering: acquisition is healthy while the profit from one deal comfortably exceeds the cost of all the enquiries it took to win that deal, conversion rates included. Comparing profit per deal against the price of a single enquiry flatters every channel and answers nothing. Everything else, including the industry average, is background noise.