We use a few cookies
Some keep the site working. The rest tell us which pages actually help people and which ones don't. Nothing optional is set until you choose, and you can change your mind whenever you like. Details in our Privacy Policy.

One Change at a Time

Changing five things at once feels productive, yet it leaves you unable to say what helped. Testing one change at a time is slower, and far more useful.

When a campaign underperforms, the instinct is to rewrite everything: new image, new headline, new audience, new page. Results might improve, and nobody learns why. The diary below follows an imaginary online furniture store through six weeks of disciplined testing. Nothing dramatic happens, which is precisely the point.

The setup. The store sells handmade oak tables. Its social ads get plenty of attention, yet few people who click go on to browse the product pages. The team agrees on one rule for the next six weeks: one change per test, and no change until the results are clear.

‍

Week 1 · Establish the baseline

‍

What changed: nothing. The existing ad runs as it is.

What they saw: strong interest in the ad, weak engagement on the product page.

Decision: the ad seems fine. Suspicion falls on what happens after the click.

‍

Week 2 · Test the landing page

‍

What changed: people are sent to a page about the dining table shown in the ad, where previously they landed on the general homepage.

What they saw: visitors stay noticeably longer and view more photos.

Decision: keep it. Matching the ad to the page clearly helps.

‍

Week 3 · Test the headline

‍

What changed: "Handmade oak tables" becomes "Built to last a lifetime, made in our workshop."

What they saw: slightly fewer clicks, yet visitors who arrive browse more deeply.

Decision: keep it. The new line attracts people who care about craftsmanship, and those are the likely buyers.

‍

Week 4 · Test the image

‍

What changed: the studio product photo is swapped for a family eating around the table.

What they saw: more clicks, with product page engagement unchanged.

Decision: keep it, with a note to test other lifestyle scenes later.

‍

Week 5 · Test the audience

‍

What changed: the ad is shown to people interested in interior design, where before it reached a broad home and garden audience.

What they saw: the numbers barely move.

Decision: no clear difference, so the simpler broad audience stays. A test with no winner still answers a question.

‍

Week 6 · Review

‍

What they know now: the page matters more than the ad; craftsmanship messaging attracts serious buyers; lifestyle images earn attention.

What they'll test next: adding delivery information to the product page, since visitors keep asking about it in messages.

What made this work

‍

Three habits kept the diary honest. Each test had a written expectation before it started. Every test ran long enough for the pattern to settle, however tempting it was to call a winner on day two. And inconclusive results were recorded with the same care as clear ones, because knowing what doesn't matter saves money too.

‍

Testing one change at a time asks for patience in week one and repays it by week six. Each finding is small, yet together they become knowledge competitors can't simply copy.

‍

Back to Blog