Campaign Index · 01 · Learn
Test distinct ideas with a clear learning question.
Start with the message, audience and decision you want the test to inform. Define concepts versus variations, inputs, comparison criteria and limitations before you spend.
A practical guide, not a benchmark
01
Plan the test before the spend.
One question
Per round
Distinct ideas
Not resized copies
Agreed in advance
How a result is judged
Planning questions, not rules that fit every account. Adjust them to your budget, category and volume.
| Component | What it is | Rule |
|---|---|---|
| Hypothesis | The one idea a variant tests | Written before anything is made |
| Concept | The idea, argument or proof being tested | Distinct, not a cosmetic change |
| Budget | A defined test budget | Set within what the business can afford to learn with |
| Decision point | When a round is read | Agreed before launch, based on budget and volume |
| Judging metric | How a result is judged | The business metric agreed in advance, not CTR alone |
| Test log | Where learnings are kept | Updated after every round |
02
How do we run a creative test?
01
Write the hypothesis
The question the test should answer, and the decision it will inform. Written down before anything is made.
02
Make distinct concepts
Same offer and landing page; each concept tests a different idea, argument or proof.
03
Isolate the budget
A defined budget, so tests do not starve what already works.
04
Judge on the agreed metric
Not clicks or likes alone. Decide the metric and the decision point before launch.
05
Log the learning
Record what the round showed and its limits, then brief the next question.
03
What did one round of testing look like?
An illustration of one round with six distinct concepts. Figures are illustrative, not client results.
Notes as ingredients
Illustrative CPA PKR 880
Founder voice-over
CPA PKR 1,140
Unboxing UGC
CPA PKR 1,020
Occasion: Eid gifting
CPA PKR 1,310
Price-first
CPA PKR 1,260
Creator reaction
CPA PKR 1,480
04
Four ways to waste a test.
01
Testing five things at once
If you change many small things without a clear idea behind them, you will not know what the round taught you.
02
Calling it on day two
Early results can swing. Agree the decision point before launch.
03
Judging on CTR
Click-bait hooks can win clicks and lose money. Judge on the business metric.
04
Not logging the learning
Without a test log, you’ll re-test the same idea in six months.
FAQ
What do teams ask about creative testing?
What is a creative testing campaign?
A defined budget used to compare distinct creative ideas against a clear question, so decisions rest on evidence rather than opinion. It does not guarantee a winner.
How much budget does a creative test need?
Enough to learn something useful within what the business can afford. We agree it before launch.
How many variants should we test?
It depends on the budget. A few distinct concepts often teach more than many small variations.
How long should a test run?
Until the agreed decision point, which depends on budget and volume. Avoid calling results too early.
Should we use Meta A/B tests or a separate campaign?
For speed, a separate ABO campaign. For a clean read on a big decision, Meta’s A/B test tool.
What should we test first?
The message: the idea or argument behind the ad. Cosmetic changes rarely teach much.
Does this work on TikTok too?
The same principles apply; platform requirements and formats differ.
Can BLITZ Campaign make the concepts?
Yes, as part of a scoped creative project. Concepts, formats and quantities are agreed in the proposal.
Next step
Want help briefing your next test?
A free preliminary review of your business, marketing and website, with suggested priorities for what to tackle next.
No obligation. We’ll contact you about the next step.