Early access

For single-location service businesses

One location. Twenty ZIP codes. A real answer on your ads.

If customers come from across a service area, your ZIP codes can do what extra markets do for a chain. Run ads in half, hold back the rest, and count the extra jobs.

Join early accessRun the free check
7730077307773147732177328773357734277349773567736377370773777738477391773987740577412774197742677433Ads on10 ZIP codesHeld back10 matched ZIPsYour locationJOBS / WEEK PER ZIPtest startsIllustrative, not a result

The problem

Your ad platforms count leads. You get paid for jobs.

Calls and form fills include people who would have booked anyway. With one location there was no fair way to tell — until your own ZIP codes became the comparison.

Before vs. after
Mixes your ads up with weather, season and luck.
Platform-reported leads
Counts people the ad touched, not people it added.
Test ZIPs vs. held-back ZIPs
Same weeks, same weather, same season. The gap is the ads.

How a ZIP-level test works

Four steps, six to eight weeks.

  1. 01

    Send weekly counts by ZIP

    Export jobs or appointments with a date and ZIP code. We count them by ZIP and week. No names, phones or street addresses.

  2. 02

    We split your area into matched halves

    ZIPs are paired by size and recent trend, so both halves looked the same before the test. That makes the comparison fair.

  3. 03

    Run ads in one half for 6–8 weeks

    Meta and Google Ads both target by ZIP. Point campaigns at the test half and exclude the rest. We size the budget first so the test can show an answer.

  4. 04

    Read the lift

    Extra jobs per week, cost per extra job and a 95% range — the method Countlift uses for chains, run on ZIP codes.

What it takes

  • You know where customers liveJobs, appointments or patients recorded with a ZIP code.
  • Enough volumeUsually 30+ jobs a week across 15+ ZIP codes. The free check sizes it.
  • A budget you can concentrateOften $2k+ a month, run in half your area during the test.
  • Willing to hold backSome ZIPs go without ads for 6–8 weeks. That’s the price of a real answer — we size it with you first.

Probably not a fit

A walk-in shop that doesn’t know where customers live, or a business with a handful of jobs a week. The free check makes it obvious.

Free guide: on/off tests for one location →

Early access

Not self-serve yet. Hands-on instead.

We’re setting up the first ZIP-level tests by hand with a small group: we size the test, split your ZIPs and read the result with you. What we learn goes into Countlift.

  • A 20-minute call to check your data and volume
  • We size the budget and hold-back before anything changes
  • You or your agency run the ads; we read the result
  • Early-access pricing when it ships
Join early access

Questions

Fair questions about ZIP-level tests.

Why not just look at sales before and after the ads?
Because weather, holidays, a competitor closing or a good month all change your numbers too. Before-and-after can’t separate those from the ads. Holding back a matched set of ZIP codes gives you a comparison that lived through the same weeks.
Won’t holding back ads cost us jobs?
Possibly a few, in the held-back ZIPs, for the length of the test. In return you learn what your ads are worth in every ZIP you serve — and whether to spend more or less from then on. We estimate the cost with you before the test starts.
What data do you need, and what don’t you?
A date and a ZIP code for each job or appointment, plus your ad spend. We never need names, phone numbers, street addresses or patient records — we turn your export into counts by ZIP and week.
We’re a walk-in business with no customer addresses.
Then a ZIP-level test won’t work. An on/off test — alternating weeks with and without ads — is the honest option. Read our free guide to measuring ads for a single location.

Find out what your ads are worth, ZIP by ZIP.

Early access · Counts by ZIP, never customer records