Brand & Content

How to test ad creative

Turn ad testing from random trial into learning that compounds: hypotheses from customers’ words, angles before details, and a correct reading of results.

NANMOW Editorial Team 5 min read

In this guide
  1. What are we actually testing?
  2. Start with a hypothesis, not a design
  3. Rules that make a test readable
  4. A simple testing matrix
  5. Creative in the Saudi market
  6. Who produces the creative, and at what pace?
  7. When to refresh creative
  8. Common mistakes
  9. In summary

In digital advertising today, platform systems handle much of the targeting and budget allocation automatically. That means one of the most important factors still in your hands is the creative itself: the idea, the image, the video, the headline and the copy.

But creative testing often turns into random trial: design several ads, launch them together, keep the “winner” — without knowing why it won. This article explains how to turn testing into learning that compounds.

What are we actually testing?

Before any test, separate two levels:

  • The idea or angle: what message does the ad deliver? For example: saving time, quality, trust, price or a real customer experience.
  • The execution: how is the idea presented? Image or video, who speaks, the first three seconds, the headline, the colors, the length of copy.

The biggest differences in results usually come from the idea, not the button color. So start by testing different angles, then refine the execution of the angle that proves itself.

Start with a hypothesis, not a design

A useful test starts with a sentence like: “We believe our customers care more about … than …, so we expect message A to perform better than message B on ….” This forces you to define what you are testing and what you will learn, whether the hypothesis succeeds or fails.

The best sources of hypotheses are customers’ own words: sales questions, reviews and pre-purchase objections. They are the same sources we use to build brand messaging, and testing is how we learn which pillar of the message persuades most.

Rules that make a test readable

Change one main element at a time

If two ads differ in idea, image and headline at once, you won’t know which change made the difference. When testing angles, keep execution as consistent as possible and change only the message.

Give each version a fair chance

Some ad systems quickly favor one ad and show the others less, so an ad looks like a “loser” before it has really been tested. Use the platform’s split-testing tools where available, or give each version enough budget.

Define the success indicator in advance

Don’t choose the indicator after seeing the results. If the goal is qualified requests, don’t judge on click-through rate alone; an ad that attracts many clicks may attract the curious rather than buyers.

Wait for enough data

A small difference between two versions over a few days may be chance. You don’t need to be a statistician, but be wary of declaring a winner on a small number of results. The fewer the results, the more time or the clearer difference you need before deciding.

IndicatorWhat it tells youIts limits
Viewing retention in the first secondsDoes the ad stop the scroll?Doesn’t tell you whether the message persuades
Click-through rateDid the ad create enough interest to click?Can rise with curiosity rather than intent
Conversion rate after the clickDoes the ad’s promise match the page and the offer?Influenced by the page, not just the ad
Cost per resultThe overall effect on your goalNeeds enough volume to settle

Read the indicators together to see where the creative succeeds and where it stumbles. An ad that stops the scroll but gets no clicks has a message problem. An ad that gets clicks but doesn’t convert usually has a gap between the promise and the page; we cover that in How to build a clear landing page.

A simple testing matrix

A practical approach is to define three or four message angles, each with two different executions. A hypothetical example for a healthy meal delivery service for employees in Riyadh:

  • Time angle: “A balanced meal before your midday meeting.”
  • Health angle: “Calories counted, without compromising on taste.”
  • Cost angle: “A weekly plan that costs less than ordering every day.”
  • Trust angle: “A licensed kitchen and menus prepared by a nutritionist.”

Each angle is executed once as a short video and once as a still image. After the test period, don’t only ask “which ad won?” but “which angle won?” That learning helps the website, content and sales, not just the ads.

Creative in the Saudi market

Some factors deserve particular attention when testing creative for a Saudi audience:

  • Language and tone: natural Arabic that feels close to the viewer differs greatly from translated copy. Local dialect may deserve a test against simple standard Arabic, depending on category and platform.
  • Faces and context: scenes, places and situations the audience recognizes build familiarity faster than generic imported imagery.
  • Seasons: messages for Ramadan, Eid and National Day need early preparation and a tone that suits the occasion, not just a decorative element added on.
  • Rules: compliance with the advertising rules in force, especially in regulated sectors such as health and finance, and with each platform’s policies.
  • Direction and design: designing Arabic creative for right-to-left reading from the start, rather than flipping an English design at the end.

Who produces the creative, and at what pace?

Continuous testing needs continuous production. A practical solution is to separate two kinds of production: core assets with higher production quality that express the brand and are used for longer, and lighter, faster test assets that try new angles. When an angle proves itself in the lighter assets, it earns a bigger production investment. That lowers the cost of experimenting and avoids spending a large production budget on an untested idea.

When to refresh creative

Even successful creative wears out over time, especially on short-video platforms where the audience sees the same ad repeatedly. Signs of fatigue include rising frequency per person along with a gradual fall in click-through rate or a rise in cost per result. Refreshing doesn’t always mean a new idea; a new execution of a proven angle may be enough.

Common mistakes

  • Testing small details before confirming the main angle.
  • Changing many elements at once, then trying to interpret the result.
  • Deciding early because of a small difference over a few days.
  • Keeping results in one person’s head instead of a written log.
  • Separating testing from the brand message, so ads become scattered voices.

In summary

  1. Start with a written hypothesis drawn from customers’ words.
  2. Test angles before details, and change one main element at a time.
  3. Define the success indicator before launch, and give each version a fair chance.
  4. Read indicators together to see where the creative stumbles.
  5. Log what you learned, and use it on the website, in content and in sales.

Organized testing turns ad spend from a cost into a source of knowledge about your customer. It is part of performance marketing when run within one system.

When you want to apply this to your brand

The principles are the same, but the right order differs from brand to brand. We start by understanding where you are today, then decide what deserves priority.