Service

Ad Creative Testing

Most ad tests change everything at once and learn nothing. We test one variable at a time — the angle first, then the hook, then the format — with organic posts finding the ideas worth paying for, and a winner rule agreed before launch.

Problem

A test that changes everything teaches nothing

New angle, new hook, new format and a new audience in the same ad: when it wins, nobody knows why, and the next ad starts from a guess. Add budgets that follow the early favorite and tests called after a day, and the account scales luck.

Scope

Four layers, tested in order

Each layer is tested with the ones above it already decided, so every result answers one question.

LayerWhat changesHeld fixedJudged by
AngleThe reason to buy — price, speed, trust, the outcomeFormat and audienceCost per lead or purchase
HookThe first seconds of a video, or the headline of a staticThe winning angleThumb-stop rate, click-through rate
FormatShort-form video, carousel, static, creator-styleThe winning angle and hookClick-through rate, cost per result
RefreshNew cuts and variations of a proven adEverything that wonFrequency and click-through rate over time
Scope

What creative testing covers

  • Organic signal first

    New angles run as organic posts to the audience the ads will target. The ones that earn attention for free are the ones that earn a test budget.

  • A test plan with a winner rule

    What is being tested, what is held fixed, the metric that decides, and how much data counts as enough — written down before anything launches.

  • Fair budgets

    Each variant gets its own fixed budget, so the platform cannot pour spend into an early favorite before the others have had a read.

  • Production built for volume

    Master templates that turn one idea into every placement and variation, with each link tagged so results report by creative rather than by campaign.

Process

Mine, pre-test, test, graduate

Paid budget is the last step, not the first: by the time an idea reaches it, the idea already has evidence behind it.

  1. 01

    Mine the angles

    Reasons to buy pulled from customer reviews, sales conversations, competitors’ ads and what already performs in your organic social — each written as one testable sentence.

  2. 02

    Pre-test organically

    The strongest angles posted organically first. Saves, shares, comments and watch time show which ones deserve paid budget.

  3. 03

    Test one variable at a time

    Angle first, then hook, then format, each on fixed budgets and judged on the metric agreed up front — only after enough data to decide.

  4. 04

    Graduate and refresh

    Winners move into scaling campaigns. When frequency climbs and click-through falls, they are refreshed with new cuts of the same proven idea.

Proof

Creative that earned its budget

−42%

Atlanta Unique Auto Sales

Paid social cost per lead, from $3.25 to $1.87, with carousel and short-form video creative tuned to in-market car shoppers, with short-form video distributed across Facebook, Instagram, TikTok, Snapchat and Google.

Read the case study →

Templates built for volume

For the same dealership, each vehicle listing becomes a Feed and a Story ad from master templates in one pass, every link UTM-tagged — so testing more variants never waits on a designer, and results report by creative.

Tag your own creative links →
Scope

The tools behind the work

Native to our workflow or run through partners — chosen for the job, not the logo.

  • Meta
  • TikTok
  • Canva
  • Google Analytics 4

Frequently asked questions

  • What is ad creative testing?

    Running variations of an ad against each other under controlled conditions to learn which element drives results. Done properly, each test changes one thing — the angle, the hook or the format — so a win tells you why it won, and the next ad starts from that answer.

  • Why test the angle before the hook or format?

    Because the angle — the reason someone should buy — carries the most weight. A better hook on the wrong message still loses. Once the message is proven, hooks and formats are refinements of something that already works.

  • How do you decide a winner?

    With a rule agreed before launch: the metric that decides, usually cost per lead or purchase, and the amount of data that counts as enough. On Meta that typically means each variant leaving the learning phase, roughly 50 optimization events in a week. Calling a test after a day is how losers get scaled.

  • Why test organically first?

    It is the cheapest signal available. An angle that earns saves, shares and watch time from your audience for free has shown it can hold attention; one that cannot is unlikely to be rescued by budget.

  • How often does creative need refreshing?

    When it fatigues, not on a calendar. Rising frequency with a falling click-through rate means the audience has seen the ad enough. The refresh keeps the proven angle and changes the execution — a new cut, a new opening, a new format.

  • Which platforms does this cover?

    The method is the same on Meta, TikTok and Snapchat, and on YouTube and Performance Max assets. The platform changes the formats and the fair-budget mechanics, not the order of the tests.

Work With Us

Find out why your best ad wins.

Book a call and we’ll look at how your creative is tested today — what changes between variants, what decides a winner, and which proven ideas are fatiguing unnoticed.