CRO

The page the ad opens,
with the version that books the job.

CRO is the test. We change one thing on the page. The ad stays in the account. The winner stays on the URL.

One change.
Then the calls.

The page is Astro, WordPress, or Shopify. One week changes the headline, the form, or the checkout. The other two wait.

Landing page

A landing page is the page the ad opens. The owner gets one job on it, the call or the form, on the client's own site.

Split test

A split test is two versions of that page. The owner keeps the version that books more of the same ad clicks.

Message match

Message match means the headline repeats the ad. The owner gets a click that still recognizes the offer on the page.

Form

The form is the fields before a lead exists. The owner gets fewer boxes that still say who is calling and from which city.

Heatmap

A heatmap shows where the visit stops on the page. The owner sees the scroll and the click in Hotjar instead of guessing.

Checkout

Checkout is the last step on Shopify. The owner gets fewer carts left full when the customer reaches payment.

We change one thing.
The winner stays on the URL.

We change one thing.

The ad stays in the account so the test is the page, not the bid. The winner stays on the URL. The other version comes down.

The test page is Astro, WordPress, or Shopify, the same stacks as the site build.

Heatmaps run in Hotjar. The call and the form go back to the ad through GA4 and server-side tracking.

Web development ships the page.

This test starts after an ad points at that page. The test waits until the page has had a week of ad clicks.

Message match is literal.

If the ad says the service and the city, the H1 says the service and the city. A homepage line about the whole company is a different page, and the ad should open the page that repeats the offer.

A headline test keeps the form.

A form test keeps the headline. A checkout test on Shopify keeps the product page, and it only runs on the payment step. A service business with a form gets a form test, not a checkout test. Two changes in the same week make the calls impossible to assign to one edit.

Scroll depth is a clue.

The winner is the version with more calls or more finished forms from the same ad and the same dates. Hotjar explains a loss. GA4 names the winner. If calls drop after the change, the old version goes back the same day.

We keep a written line of what changed and the date.

The ad budget stays where Marketing left it. This page does not move bids. A tie, same calls on both versions, stays up another week on the same ad, the same URL, and the same two versions.

The test URL is a path on the business's own domain.

The ad does not open a second host. When the winner stays, the old line remains on the written list so we can put it back the same day if the calls fall.

On Astro the test is a second version of that page in the same project.

On WordPress with Valor the test is a second template, not a second pile of plugins. On Shopify the test is a second template in the same theme. The domain the customer typed stays the domain.

The note at the end of the hold is four lines: the line that changed, the dates, the calls on each version, and which URL stays.

Both versions use the same ad URL, so the auction is not what moved.

GA4 and the server-side hit say which version got the call. Hotjar says whether the visit reached the form.

Owners in The Woodlands, Houston, Conroe, Spring, and Magnolia get the test on the page their ads already open.

The ads themselves sit on Marketing. The follow-up note sits on Email and SMS.

Read the page.
Then change one line.

  1. 01

    Read

    The ad, the page, the form, and the calls from the last 30 days. Name the leak before writing a new line.

    Week 1
  2. 02

    Change

    One headline, one form, or one offer. The other two stay put until this test has a winner.

    Week 2
  3. 03

    Hold

    Both versions take clicks from the same ad. We pick the winner from calls and forms, not from a hunch.

    Weeks 3-4
  4. 04

    Keep

    The winner stays on the URL. The next test starts on the next leak, one change again.

    After
Bring the ad and the page. We will say which one is leaking. Begin Private Audit