Spend against every new client
Total ad spend against all new clients, by month and by location. The one number no channel can claim more than its share of.
Tracking sees clicks and promo codes, so it gives brand search the credit for people who already knew your name, and it can't see a billboard at all. New clients remember the billboard they drive past and forget the ads they scrolled by. Total spend against every new client shows the whole, but not the parts. Read together, each fills in what the others leave out.
| Channel | Tracking says | Clients say | What it means |
|---|---|---|---|
| Google non-brand search | $70 | $97 | They agree. A safe place for more budget. |
| Google brand search | $21 | $80 | Tracking also counts people who were already looking for you. A test shows what it adds. |
| Billboards | Can't see | $188 | Only clients see it, and they remember what they drive past. A test shows what it adds. |
| Meta ads | $292 | $143 | Tracking sees the clicks; clients also remember the ads they only saw. Don't cut it on clicks alone. |
Total ad spend against all new clients, by month and by location. The one number no channel can claim more than its share of.
Clicks, booking links and promo codes, by channel. Strong on search. It can't see billboards, or ads people saw and didn't click.
Asked at sign-up: every channel that played a part, then the one that decided. Credit is shared, so a person moved by three channels still counts once.
Pause or raise one channel at half your locations, picked by coin flip in pairs of similar size. The other half show what would have happened anyway.
Fake tests on your own history show how much new clients bounce with nothing changed. A test is planned only when its two possible answers land further apart than that.
Before anything changes, we write down the result that would move the budget. The test is read once, at the end, against that line.
How many new clients a test could cost, what it saves or spends, and what could spoil it, signed off before it starts.
Some questions can't be settled at your size. The plan says so, and gives the range to plan on instead.
Ad spend from each ad account and your books, and new clients from your booking system, by location.
If your sign-up form doesn't ask how people heard about you the same way every time, we set the question up with you.
Each month: the three signals by location, what changed, and the next test worth running.
It runs on the same connections as the Data Engine. Read when a marketing mix model fits a multi-location business.
It starts from the same idea, total spend against total results. But a full marketing mix model needs spend that changes on its own, not with the season, and ads that bring in a large share of new clients. Most chains of 5 to 50 locations don't have that yet, so we start with the three signals and tests, and add a model only when your data can support one. Read the guide.
Any channel you spend on: search, Meta, TikTok, Yelp, billboards, mailers, events. Tracking sees some of them. What new clients say covers the rest, and tests settle the biggest questions your number of locations allows.
No. Your team or agency keeps running them. We only measure, so nobody grades their own work.
No. Every number comes from code under written rules, and you can trace each one back to the sign-ups, survey answers and spend behind it.
A setup fee, then a monthly fee, quoted once we've seen your channels and locations.
Tell us your channels and how many locations you run. We'll reply with what we'd measure first.