Questions this report answers
- Will the ad be noticed, and will viewers credit it to our brand rather than a competitor?
- Which message do people take away, and is it the one in the brief?
- Does the ad raise consideration or preference among the target group?
- Which of two or three creative routes should go into production?
- Which scenes, lines or claims confuse or irritate viewers?
- After the campaign, who recalls it, and did brand measures move?
- Which headline, visual or call to action works best online?
When you need it
A pre-test pays off when a decision is expensive to undo: choosing between creative routes at storyboard or animatic stage, signing off a spot before a national TV or video schedule, adapting an existing ad for a new country where humour or casting may not travel, or using a claim in a regulated category such as over-the-counter medicines. A post-test belongs at the end of every significant campaign, because its findings shape the next brief.
Digital media have changed the mix rather than removed the need. Eurostat reports that 32.6% of EU-27 enterprises with ten or more employees paid to advertise on the internet in 2024, with 41.5% in the Netherlands and 25.6% in Bulgaria. Online campaigns are often optimised through platform split tests, which show which version earns more clicks but not why, and not what happened to the brand among the many people who saw the ad and never clicked.
Full survey-based testing is not worth it for small, short-lived digital bursts, where a live split test answers the question faster, or for creative that changes daily. If the problem is the proposition rather than the execution, go back to concept testing first.
Sample table of contents
- Executive summary
the verdict on each creative and the changes recommended - Objectives and benchmarks
what the ad must achieve and how success is judged - Methodology and sample
exposure design, test and control cells, quotas - Attention and branding
standout, brand recall and correct attribution - Message take-out
what viewers think the ad says, coded against the brief - Persuasion
change in consideration or preference against the control cell - Emotional response
likeability and scene-level reactions - Creative diagnostics
scenes, lines and claims that help or hinder - Post-campaign results
ad recall and brand measures against the pre-wave - Recommendations
edits, cut-downs for social and points for the next brief
Methodology and sources
Pre-test
Respondents from the target group see the ad in a realistic setting, such as a reel of other ads or a mocked-up social feed, so that attention is not inflated by a blank screen. Unprompted questions come first: which brands and ads they remember. Prompted questions then cover branding, message, likeability, believability and consideration. A matched control cell that sees no ad, or the current one, lets us read persuasion as a difference rather than an absolute score. A moment-by-moment slider shows where interest rises and falls, and open answers are coded. This is what an advertising pre-test adds that a platform metric cannot.
Post-test
An advertising post-test interviews the target group after the campaign, showing stills with the brand masked to measure genuine recall, and compares brand awareness and consideration with a pre-campaign wave or with people who were not exposed.
Digital creatives and quality
For platform split tests we help define the variants and read the results; our A/B test significance calculator shows whether a difference in click-through is real or noise. Every survey uses device quotas, attention checks and removal of speeders.
What you receive
- Report in PDF with a scorecard for each creative and recommended edits
- Scene-by-scene diagnostic charts
- Message take-out coded against the creative brief
- Excel tables by segment, country and media exposure
- Topline within days of fieldwork closing
- Debrief workshop with your brand and agency teams
How the project runs
- Briefing — 2–3 days to agree objectives, target group and success benchmarks
- Questionnaire and test environment — 3–5 days to build the ad reel or feed mock-up
- Fieldwork — 4–7 days for a pre-test; a post-test runs in the fortnight after the campaign ends
- Analysis and coding — 4–6 days
- Report and debrief — 2–3 days
Worked example
Example: a hypothetical brand of non-drowsy hay-fever tablets, sold through pharmacy chains in Poland and Romania, has two online video routes for the spring season. Route A uses humour built around a sneezing wedding photographer; route B is a plain demonstration of the non-drowsy benefit.
A monadic pre-test with hay-fever sufferers in both countries, plus a control cell, shows that route A is better liked and better remembered, but many viewers attribute it to the market leader. Route B is less entertaining, yet branding is strong, the non-drowsy message is played back clearly and consideration rises against the control. The wedding joke also lands less well in Romania.
The brand runs route B as the main spot, borrows the opening scene of route A as a hook for short social cut-downs, split-tests two thumbnails on the platform and books a pre-campaign wave plus a post-test at the end of the season.
Common mistakes and how to avoid them
- Showing the ad alone on a blank screen. Attention is overstated; embed it among other ads or in a feed.
- Judging by likeability. A well-liked ad can still be poorly branded, so read attribution and message take-out alongside it.
- Skipping the control cell. Without an unexposed group, persuasion cannot be separated from existing brand strength.
- Writing a new questionnaire for every test. Standard measures make results comparable across creatives and years.
- Reading click-through as a brand effect. Clicks reflect a small, active group; recall and consideration need a survey.
Who it suits
| Suitable for | Why |
|---|---|
| Brands committing to a large TV or video schedule | Fixes weak branding before the airtime is booked |
| Agencies choosing between creative routes | Gives an independent read on which route to produce |
| OTC and pharmacy brands with claim-led ads | Checks that the claim is understood and believed |
| Retailers with seasonal campaigns | The post-test shows what was recalled and feeds the next brief |
| International brands adapting ads for new markets | Reveals humour and cultural cues that do not travel |
Frequently asked questions
How long does an ad test take?
Two to four weeks is typical from briefing to debrief. Pre-test fieldwork usually takes under a week; building the test environment and coding open answers take the rest. A post-test is scheduled around your campaign dates, ideally with a pre-wave before launch.
What affects the cost of ad testing?
The number of creatives and cells, including the control, the size and incidence of the target group, the number of countries and languages, and extras such as scene-level diagnostics or a pre-campaign wave. A short brief with your media dates is enough for us to prepare a quote.
Can you test an ad before it is finished?
Yes. Storyboards, animatics and rough cuts can all be tested, which is often the best moment to choose between routes. Unfinished material tends to be rated differently from finished film, so we compare like with like and suggest a final check on the edited spot when the stakes are high.
How is this different from A/B testing on social platforms?
Platform tests measure real behaviour, but only clicks and views among people the algorithm chose. They say nothing about why one version won or about brand effects among non-clickers. Survey-based testing supplies those diagnostics, and the two work best together, with surveys for direction and live tests for fine-tuning.
How do you tell whether a campaign actually worked?
By comparing recall, awareness and consideration after the campaign with a pre-campaign wave or with an unexposed group. If you need to follow these measures over several campaigns, the post-test can become a wave of continuous brand health tracking, with digital market analysis adding search and social signals.
Is our unreleased advertising kept confidential?
Creatives are streamed from a secure player and cannot be downloaded, respondents agree not to share what they see, and the survey never names your agency or the launch date. Respondent data is processed in line with the GDPR and the ICC/ESOMAR International Code.
Related reports
Brand health tracking
Brand health tracking is a repeated survey that measures how many people know your brand, consider it, prefer it and…
Read →Concept & product testing
Concept testing measures how target buyers react to a new product idea, recipe, pack design or brand name before you…
Read →Digital market analysis
Digital market analysis studies a market through the traces people leave online: what they search for, what they say on…
Read →Focus groups
A focus group is a moderated discussion in which six to eight carefully recruited participants explore a topic, product…
Read →Sources
Updated: 5 Nov 2026