How cost per conversation is calculated
Take everything the campaign spent over a period and divide it by the number of conversations started in the same period. Meta reports the result directly, so most people meet it as a column rather than as a calculation.
Because the denominator is threads rather than enquiries, the figure moves for reasons that have nothing to do with sales. A cheaper audience, a lighter auction, or an advert that makes people curious enough to tap will all bring it down. So will a greeting that invites people to say hello without asking for anything. None of those changes necessarily produces a single extra customer.
Why cost per conversation matters
It is the fastest signal a messaging campaign gives you. Threads open within minutes of a launch, so you can tell early whether the audience and creative are working, long before there is a sales outcome to read.
It is also the honest way to compare messaging campaigns with each other. Two audiences, two offers or two openings can be judged on the same terms, and a difference between them usually points at something you can act on: the wrong destination app, an offer that is too vague, or an audience that was never in the market.
Common mistakes with cost per conversation
Setting it against cost per lead is the one that misleads businesses most. Opening a chat costs the person a tap; filling in a form costs them their details. The same budget will always buy more threads than leads, so messaging looks cheaper on paper while the two numbers are measuring different levels of commitment. Switching channels on that comparison alone is a decision made on an illusion.
The second mistake is driving it down deliberately. It falls when conversations get easier to start and lighter in substance, which is the opposite of what a sales team needs. The third is reacting to weekly swings — auctions move, festivals and paydays move demand, and a single bad week is rarely a reason to change anything.
How to act on it
Build your own version of the figure and keep both. Divide spend by the conversations your team judged to be real enquiries, and divide it again by the ones that became customers. Those two numbers are the ones worth defending in a budget meeting; the reported column is an early indicator.
Then use the reported figure the way it is good for: comparing like with like inside the same account, over a period long enough to mean something. When it rises, look at what changed — the audience, the creative, the destination, the season — before touching the budget. When it falls, check that the extra threads are still worth answering, because a bigger inbox full of nothing is an expensive kind of cheap. The way WhatsApp and Messenger ads are structured should follow what your team can genuinely handle.
