CPA / CPL Calculator

What this tool does

CM8-370 works out what each campaign pays for a lead and for a customer, and then breaks that cost apart into the three things that produce it: what you pay for attention, how many people click, and how many of those go on to convert.

Everything runs inside this single file — no account, no upload, no network request of any kind.

The chain, and why it matters

Each cost is the one before it divided by a conversion rate. That is the whole structure, and it is why cost per customer is the least useful number to look at on its own: by the time you see it, three separate things have multiplied together and you cannot tell which one moved.

A high cost has only two causes

Either you are paying too much for the click, or too few of them convert. That is the entire space, and the fixes have nothing in common:

• Expensive clicks — a bidding, targeting or creative problem. Fixed in the platform: better keywords, tighter audiences, higher relevance, different placements.
• Poor conversion — a landing page, offer or product problem. Fixed on your own site, and often much cheaper. Nothing you do in the ad account will help.

The drivers chart puts cost per click and conversion rate on the same row for exactly this reason. In the sample, competitor-term search has by far the most expensive clicks and one of the best costs per customer, because the people clicking are already shopping.

Qualified leads

The most useful field in the tool, and the one most accounts do not fill in. A campaign producing cheap leads that are all unqualified is producing nothing at all — it is converting your money into work for the sales team.

The sample makes the point deliberately. Paid social prospecting shortened its lead form from nine fields to four: lead volume doubled, cost per lead halved, and the qualified share fell from 42% to 22%. On cost per lead it looks like a triumph. On cost per qualified lead it is roughly flat, and the sales team is now working twice as many bad leads.

Any campaign where fewer than a third of leads are worth calling is flagged, whatever its cost per lead says.

Too small to judge

A conversion rate from a handful of outcomes is noise. Three customers from four thousand clicks tells you almost nothing: one more or one fewer moves the rate by a third.

Campaigns below the threshold are shown in the table with all their figures, marked too small to judge, and left out of the ranked charts entirely. Thirty outcomes is a low bar and even that is thin; it is enough to stop the worst decisions.

Setting the ceiling

The most you can afford to pay for a customer comes from margin and lifetime value, not from what you happened to pay last year. Roughly: the gross profit a customer produces over the period you are willing to wait, divided by the return you need on acquisition spend.

Anchoring on last year's cost per acquisition guarantees you will never find out whether you could profitably pay more — which is usually the constraint on growth rather than the cost itself.

Making the two ceilings agree

The two targets have to be consistent, and in most accounts they are not: somebody set a lead target years ago and a CPA target last quarter, and they imply different conversion rates.

The tile shows what your actual lead-to-customer rate implies your lead ceiling should be. If it comes out well below the ceiling you set, your lead target is too generous and campaigns are being praised for hitting it while missing the one that matters.

Sales time

Where leads get called, the media cost is only part of the story. Set the hours per lead worked and the cost of that hour, and the tool adds it:

Where qualified counts exist it uses those, since unqualified leads are not worked for long. This is the figure that reveals a cheap-lead channel to be an expensive customer channel — a campaign producing eleven hundred leads at fourteen each is consuming a great deal of somebody's week.

This is not customer acquisition cost

Everything here is media cost per outcome, plus optionally the sales time. It does not include agency fees, creative production, tools, or the marketing team's own hours.

That is deliberate: this tool is for optimising campaigns against each other, where those shared costs would only add noise. For the fully loaded figure — the one that belongs in a business case — use an acquisition cost register, which counts everything.

Reading the diagnosis table

For each campaign, the tool compares all three steps against the best in your own account and names the one furthest behind. The last column shows what the cost per customer would be if only that step matched the best.

It is a prompt, not a promise. The best campaign is usually best for reasons that do not transfer — brand search converts well because those people already decided. Use it to decide where to look, then go and look.

What this does not do

• It counts what the platform reports, with all the attribution questions that carries.
• It has no view on whether a lead is real. A campaign generating fake form fills will look excellent until somebody records the qualified count.
• It does not know about the time between a click and a sale. In a long cycle, this month's spend and this month's customers are not related.
• It does not measure value. A campaign with twice the cost per customer may be bringing customers worth three times as much.
• It has no view on diminishing returns: cost per customer rises as a campaign scales, and nothing here predicts by how much.

Printing and sharing

The Report tab prints the tiles, charts and both tables with a title block you fill in. The chain table is the working document for an optimisation session — every rate and every cost, one line per campaign.

Saving your work

Campaigns are held in this browser, on this computer, and stay there between visits. Use the backup button to write a JSON file you control.

Accuracy & disclaimer

Every figure here is one you entered from a reporting system with its own definitions. The tool divides and compares; it has no view on whether a campaign should run or whether a lead was worth having.