Can you hold the cost per customer you promised?
A four step funnel where every step is a rate and the answer is a ratio. The cost per acquisition the plan calculates is not the average cost per acquisition you will get, and the gap between them is the size of the promise you should not make.
Words on this sheet
- Cost per acquisition: What you spend, on average, to win one new customer.
Marketing Advanced Monte Carlo Pro engine
After you install, this is the model to open.
Will the Campaign Hit the Cost Per Acquisition?
- In your spreadsheet, click the Sortia icon in the strip of icons down the right-hand edge. No strip? Click the arrow at the bottom-right to open it. You can also use Extensions, then Sortia, then Open Sortia.
- Click Start from a template and put that name in the search box.
- Pick the card with that name and click Load this template. It arrives on a new tab with real numbers already in it.
This one runs on a Pro engine, and every free install includes five full-quality runs on your own numbers, shared across all five Pro engines rather than five for each. After that, Pro is $199/year.
The answer
- The plan says
- $230.45 per customer, and it goes in the contract
- The simulation says
- $243 median, with a mean of $256
- Odds of missing $240
- 51% a coin flip on four uncertain rates
- A number you can hold
- $302 P75: kept three quarters in four
Four rates and a budget. At the numbers in the plan, $60,000 buys 4.29 million impressions, 38,571 clicks, 3,471 leads and 260 customers, which is $230.45 each, and that is the number that ends up in a contract. Click Run with all four rates on ranges and the first result is the one to understand before anything else. The average cost per acquisition is not $230.45.
It is about $257. The median is about $243. Neither of them is the number the plan calculated, and the reason is not a modeling error: cost per acquisition is a budget divided by a product of four uncertain rates, and dividing by an uncertain number is not the same as dividing by its average. The bad quarters land much further from the plan than the good quarters do, so the average sits above the plan and the median sits just above it.
The full picture: P5 about $151, P25 about $198, median about $243, P75 about $302, P90 about $365, P95 about $407. The promised $240 is missed in about 52 of every 100 runs. If you have to write a number into a contract, the P75 near $302 is the number you can hold in 75 of every 100 runs, and the plan number is the number you can hold in about 50.
The tornado ranks click through rate first, landing page conversion second, lead to customer rate third and the cost per thousand impressions last, which is the reverse of the order in which media buyers argue about them. The second run is a real commercial decision rather than a modeling exercise. Buy the same campaign on a cost per click basis at $1.56, which is what this plan implies, and the platform carries the delivery risk instead of you: put the cost per click where the cost per thousand impressions sits, make clicks the budget divided by it, drop the impressions and click through rate rows, and rerun with only the two remaining rates on ranges.
The median barely moves, from about $243 to about $234, but the P90 comes in from about $365 to about $309 and the miss rate falls from 0.52 to 0.45. That $56 of P90 is what the impression buy is costing you in promise keeping, and it is what you should be willing to give up in a higher cost per click. One thing the model gets wrong on purpose: it treats the four rates as independent, and they are not.
Creative that lifts click through rate usually lowers landing page conversion, because it brings a broader audience. If your own data shows that, add a correlation pair on those two at about negative 0.4 in the Correlations panel and rerun: the median stays put at about $245, the P5 rises from about $151 to about $161 and the P95 falls from about $407 to about $377.
The tails come in and the middle does not move, which is what a correlation does and what people expect it to do to the average. To make it yours, take all four rates from the last two quarters of the same account rather than from a benchmark, and put the number you are being asked to promise into the promised cost per acquisition row.
The model
It arrives on a tab called Template: Will the Campaign Hit the Cost Per Acquisition:
| Media budget ($) | 60000 |
| Cost per thousand impressions ($) | 14 |
| Impressions | 4,285,714.3 |
| Click through rate | 0.009 |
| Clicks | 38,571.4 |
| Landing page conversion to a lead | 0.09 |
| Leads | 3,471.4 |
| Lead to customer rate | 0.075 |
| Customers | 260.4 |
| Cost per acquisition ($) | 230.5 |
| Cost per acquisition promised to the client ($) | 240 |
| Misses the promised cost per acquisition (1 = yes) | 0 |
Once it is in your sheet
- The model arrives with real numbers in it and runs as it stands, so you can press the button first and understand it second.
- Change the numbers to yours. The sheet marks which cells are inputs and which hold formulas, and most labels carry a note explaining the row.
- Press the run button at the bottom of the panel. It is labeled for the tool you are in, and the result lands on its own tab, with a written reading of it beside the figures.
Never used Google Sheets? Start here goes the whole way, in seven steps, and assumes nothing.
Next question
- What would have to change for this channel to pay?What Conversion Rate Makes This Channel Work?
- The winner got more opens. Did it get more clicks?Which Subject Line Won?
- Is a pitch you probably lose still worth entering?Should We Pitch for This Account?
- Does the price cut earn back the margin it gives away?Will the Promotion Pay For Itself?
- How many invitations does sixty replies take?Pull the People to Survey
- How much is riding on your one big channel?What If the Ad Platform Changes the Rules?
Every model like this one, and the method behind them: Monte Carlo simulation.