Most companies can tell you what they spent last quarter. Far fewer can say what a customer is actually worth, what each channel costs to acquire one, or which of the channels they have written off were never really tested.
The Growth Diagnostic answers those three questions from your own data and ends with a ranked list of what to test next. Fixed scope, fixed fee, results presented within three weeks of start. It takes weeks rather than a quarter, and it usually changes where the next dollar goes.
If you run a subscription business
Margin LTV comes off ARPU, average churn and gross margin, and the cost-per-customer ceiling follows from it. Payback is reported alongside the LTV:CAC ratio, because a business can clear 3:1 and still be unable to fund the growth that implies. When customers churn matters as much as how many — losing them in the first few months lowers the ceiling in a way an average churn rate hides.
If you run an e-commerce business
The ceiling comes off contribution margin per order after cost of goods, shipping, fulfillment, processing and returns, computed on net revenue so discounts are already inside the number. How margin builds across the first, second and third order matters too: the total sets the ceiling, and the timing sets how fast the cash comes back and how much of that ceiling rests on repeat behavior that has not happened yet.
Two things. The first is the assessment: a working model built from your own customer or order file, not a slide deck. What a customer is worth, the acquisition ceiling that follows from it, and your current cost per customer channel by channel — the two numbers set side by side, which is where the gaps become obvious.
The second is a prioritized test queue: a ranked list of tests you can run to find out how to grow sales and lower your cost of acquisition, ordered by what each one costs and how quickly it gives you an answer. Where the numbers point to a pricing or offer change rather than a new channel, that goes on the list too. Channels where an earlier test was too small or too short to prove anything go back on the list — those are untested rather than failed, and they are usually where the opportunity is.
You keep the model. It is built to be used after the engagement ends, not presented once and filed.
1. What I need from you
Your customer or order file with the acquisition source, ad platform exports, total marketing spend by month, gross margin, and a conversation about what you have already tried. Approximate figures are fine and no cleanup is required. Where a piece of it does not exist, that is itself a finding.
2. How it runs
A kickoff call covers your test history, your constraints, and anything you are unwilling to try. From there the work sits on my side — you send what you have and I come back where figures disagree. Three weeks is the outside date, and it assumes normal turnaround on the data rather than anyone clearing their calendar.
3. The readout
A live readout with the working model open. You get the model and the queue, not a presentation. There is also a version of this where the output is that a growth test is not the right next step — because the economics need establishing first, or the lever that matters is one the business will not pull. In that case I say so rather than design a test around it.
4. What it costs
A fixed fee, agreed before the work starts, with no retainer commitment attached. It credits in full against the test design and execution work if you want me to take that on. I will give you the number on the first call, once I know what shape your business is in.