WhatRunsIndiaMany small brands run "creative tests" that are really decoration tests. Orange button versus green button. Headline in bold versus regular. After two weeks the results are a coin toss, and the team concludes that testing does not work for them. The problem is not testing. It is what is being tested.
Why do small tweaks rarely show a winner?
To tell two ads apart, the difference in results has to be bigger than the normal randomness in your data. A button colour might move results by a few percent at most. Detecting that reliably needs a lot of purchases, far more than a brand spending a few thousand rupees a day will collect in a week.
A big change, like a different scene or a different reason to buy, can move results by much more. That is a difference a small budget can actually see.
What counts as a concept?
A concept is a different answer to the question "why would someone stop and buy this?" Some examples for an ethnic wear brand:
- Occasion: the kurta at a Diwali puja versus at the office.
- Person: a college student versus a working mother versus a grandmother.
- Proof: a close-up of the hand-block print versus a crowd of happy customers.
- Offer: free shipping versus a festive bundle versus "new drop".
- Setting: a heritage courtyard versus a clean studio versus a city street.
Each of these changes what the ad says, not just how it looks.
How should a concept test be set up?
Keep it simple:
- Pick one product that already sells, so you are testing the creative and not the product.
- Make 3 to 5 concepts, each clearly different.
- Keep everything else equal: same ad set, same budget, same offer unless the offer is the concept.
- Judge on cost per purchase, with click-through rate as an early signal only.
- Give it time: about a week, or until each concept has a meaningful number of purchases.
What happens after a concept wins?
Now small variations become worth making. Take the winning concept and create versions with different crops, models, headlines and formats. These keep the idea fresh as it tires, and some will outperform the original. Our guide to creative fatigue explains when to swap them in.
This is also when button colours and fonts can matter, at scale, on a winning concept, with enough volume to measure. Not before.
How do you produce enough concepts?
The real barrier is production. Five concepts for one product used to mean five different shoots. WhatRunsIndia lets you put the same product into different scenes, on different models and in different ad layouts from your existing product photo, so a concept test costs minutes rather than a shoot.
FAQ
What is a creative concept in advertising?
A concept is the core idea of an ad: who it shows, where, and why someone should buy. Changing the concept changes the message, not just the design.
How many creative concepts should I test at once?
Three to five clearly different concepts for one product is a practical number for a small budget. More than that spreads the spend too thin.
Is A/B testing button colours a waste of time?
On small budgets, usually yes, because the differences are too small to detect. Test big concepts first and save fine-tuning for winning ads with high spend.