The number we will not print
We publish no cost-per-lead benchmark of our own, and we would encourage you to distrust the ones you are shown. The figures circulating in this market come from vendor content, describe different things, and contradict each other. Some count a portal click. Some count a form submission. Some count a phone call that rang out. A precise-sounding number built on an undefined denominator is worse than no number, because it invites you to compare your desk against something that was never measuring your desk.
So here is your own arithmetic instead. It takes about twenty minutes and a portal invoice.
The first number, which most desks already have
Take last quarter's portal invoice, the whole thing, including credit top-ups and any feature or premium placements. Divide it by the number of enquiries it produced.
That is your cost per enquiry. Most brokerages either know this or can get it inside an hour, and it is the number that gets quoted in management meetings.
It is also the flattering one.
The second number, which almost nobody has
Run the same division again, with a different denominator. Take the same invoice, and divide it by the number of enquiries that actually got into a conversation. Not the ones that were assigned. Not the ones an agent opened. The ones where a real exchange happened with a human being who replied.
That is what you are genuinely paying per conversation, and it is the figure that should drive decisions. On most desks it is several times the first number, and the gap between the two is the entire size of the opportunity sitting inside your existing spend.
Nobody outside your building can tell you what that gap is. Not us, not a portal account manager, and not a supplier with a case study. It depends on your coverage hours, your routing, your follow-up discipline and your market, and all four of those are yours.
Why the second number is the one that moves
Portal packages in this market are sold against listings and placement. You buy a subscription tier or a pool of credits, and you spend them putting stock in front of people through standard listings, featured slots and premium positions. The details differ between platforms and change between contract years, so check your own agreement rather than trusting a summary of it, including this one.
What does not differ is the shape. You are billed for placement, and nothing in that model refunds you for the enquiry that arrived at 21:40 and got answered on Sunday. The money left your account when the listing went up. What happened afterwards is entirely on your side of the wall, and it is invisible on every dashboard the portal gives you.
That is not a criticism of the portals. They are selling placement and they deliver placement. It is a statement about where your leverage is. If your cost per enquiry is fixed by a contract you signed at the start of the quarter, then the only variable left is what proportion of those enquiries turn into conversations. That number is not fixed, it is not visible, and on most desks nobody owns it.
The third number, if you want to go further
Split the conversation figure by hour of arrival.
The HBR audit that first put a size on this problem found the causes were structural: enquiries pulled from a CRM in daily batches rather than continuously, and routing rules built around fairness rather than availability. Both of those show up as a clean pattern when you split by hour. Office-hours enquiries convert into conversations at one rate, and evening and weekend enquiries at a much lower one.
If your split is flat, your problem is qualification quality and you should ignore anything anyone tells you about response speed. If your split falls off a cliff after 18:00, you have a coverage problem, and no amount of training, targets or CRM discipline will fix a coverage problem, because the agent you are training was asleep.
What to do with a supplier who quotes you a number
Ask them three questions.
What counted as a lead in that figure, exactly. Whose data was it, and can you see the sample size. And was the comparison group doing the same thing as your desk, in the same market, at the same time.
If the answers are vague, the number is decoration. That test applies to us as much as to anyone else, which is why there is no cost-per-lead statistic anywhere on our site and why this article ends without one.
What we would actually claim
Eric picks up the enquiry when it arrives, at any hour, and runs the qualifying conversation on WhatsApp before an agent is involved. The claim there is about coverage and about the record it produces, not about a conversion multiple we have not measured.
The reason we care about your second number is simpler than a sales argument. If the gap between your cost per enquiry and your cost per conversation is small, your desk is already answering its inbound and we have very little to offer you. If it is large, you have already paid for those enquiries. That is the whole case, and you can check it yourself before you ever speak to us.