A few weeks from now, on a call you haven't booked yet, someone is going to ask you where your customers came from. Not what you sell, not what you can spend — where the people who have already paid you heard about the business in the first place. It sounds like warm-up before the real conversation. It is the real conversation, and the rest of the engagement gets built on whatever you say next.
Most owners answer with a story. Word of mouth, mostly. A neighbor posted something. A few came from Instagram, probably. All of that can be true and still be useless, because a story can't be compared to anything. The person you hire is going to spend your money by comparison: this channel against that one, this week against last, a paid customer against one who walked in free. Without a record, the first comparison they can make is months away, and you pay for the wait.
Why they ask that before anything else
Every decision a marketer makes for you is a decision about what a customer is allowed to cost. That ceiling comes from your history — what people buy, how often they come back, how long they take to decide. A new business has thin history, which is fine; what isn't fine is having none of it written down. When nobody can name the ceiling, the default becomes spend until it feels uncomfortable, then stop. That is how a launch budget disappears with nothing learned from it. If you'd rather work the problem from the customer end first, the demand side of this is covered in how to get clients for a new business.
Who else is typing this phrase, and when
«New business marketing» is not a big search. The whole cluster runs about 240 a month, and that total is 5 distinct queries merged together — people circling the same worry in slightly different words rather than a crowd shopping for one thing. What's notable is the price attached to it: top-of-page bids on this cluster reach $55.25. Somebody is willing to pay that much for one click from an owner in your exact position. Treat that as information about the advice market, not about yours — and as a reminder that bidding blind is expensive in any category where the buyer is valuable.
The interest also swings hard by month. Searches for the phrase hit 320 in September and fell to 50 in December. Owners plan in the fall and stop reading in the holidays, which means the quiet stretch is when the unglamorous setup work actually gets done. If you are reading this in a busy month, the setup still comes first; it just competes with more noise.
The sheet that answers the question
You don't need software for this yet. You need one sheet, filled in every time money changes hands, with columns that a stranger could read and act on. Four columns cover the question you'll be asked.
| What you write down | Where it comes from | What it lets a hire decide |
|---|---|---|
| How they found you, in their own words | Ask it at booking or at checkout | Which channel gets tested first |
| Date of first contact, date they paid | Your inbox, calendar, invoices | How long a sale actually takes |
| What they bought and what they paid | Invoices | What a new customer is allowed to cost |
| Whether they ever came back | Repeat invoices or bookings | Whether an expensive click still pays |
Where the record quietly stops working
It fails in predictable places. Staff stop asking on busy days, so the record skews toward slow ones. Phone calls never get logged, so the channel that drives them looks dead. The website form starts emailing you and nothing else, so no analytics tool ever sees a lead. Someone redesigns the booking page and the tracking that was working goes silent without an error message. None of these announce themselves — you just end up with a clean-looking sheet that describes a business you don't have. Checking that the booking or form actually registers as a conversion, in the tool you'll eventually judge ads with, is worth more than a month of posting.
What a hire can and cannot do with it
A freelancer, a small shop, or an analytics-minded agency can wire the measurement properly, get the ad account structured so the spend maps to something, and read the result back to you in plain terms. That is real work and it is reasonable to pay for it. What nobody can do is invent the history — if you hand over a story, the first months of any engagement go to building the record you could have built yourself for free, at agency rates. So bring the sheet to the first call, however short it is. The gap between «word of mouth, mostly» and a dated list of how people found you is the difference between buying a plan and buying a guess.