They loved the ad.
What did they think
you were selling?
A campaign can be entertaining, beautifully made and understood in a way its creators never intended. Before rewriting the headline, find out which offer the audience actually saw.
Imagine a campaign review where everyone has a favorite part of the video. The opening line. The music. The dog. Especially the dog.
Then someone asks what a prospective customer would say the company actually offers. The room gets less specific. “Freedom,” perhaps. “Confidence.” “A better way to work.” These may be perfectly good ambitions. They are also available from a remarkable number of businesses, including the company that sells the dog food.
I would want another answer before approving the creative: what should someone understand after seeing it, and what evidence tells us that meaning arrived?
This is a narrower question than whether the campaign performed. A person can understand an offer and decline it. Someone else can click because they misunderstood it. Enjoyment, memory, comprehension and purchase are related things to investigate, but none gives you a complete reading of the others.
My concern here is the moment a piece of creative becomes an interpretation in somebody else’s head. That interpretation may be reasonable, even when it is wrong about your business.
Read this before I explain it.
This is an original, fictional advertisement for an invented company. There are no customer findings or campaign results behind it. Read the ad, decide what you believe it offers, then open the explanation.
Get your
Tuesday back.
Fewer tabs. Less chasing.
A business that runs without you.
Reveal the intended offer and inspect the mismatch
For this example, Tuesday sells a fixed-scope consulting engagement to map a small company’s recurring administrative work and recommend which tasks to change, delegate or automate. It does not sell software, supply assistants or run the company. Its first conversation is a fit discussion. The engagement itself is paid.
The creative supplies a benefit and a mood. It leaves the category largely to the viewer. “Fewer tabs” could suggest software. “Less chasing” could suggest an assistant. “A business that runs without you” could suggest outsourced management or a promise of far more independence than the engagement delivers.
Those are plausible readings of these words. They are editorial hypotheses, not recorded audience responses. We would need research with appropriate prospective buyers to learn which interpretations actually occur.
The fictional offer exists only to examine the creative. It is not a Paper Boat Media package or a claim about a real advertiser.
The audience may have understood every word.
There is nothing difficult about “get your Tuesday back.” The problem is not vocabulary. It is the range of businesses the sentence could reasonably describe.
A company’s employees supply the missing context automatically. They know what the product does, how the service works, what the sales team explains and which customers fit. An unfamiliar viewer has the material in front of them. That can include the account name, image, caption, headline, button and surrounding placement. It does not include the kickoff meeting.
When reviewing an ad, I want to separate what the creative states from what the business hopes the audience will infer. An inference is not necessarily a failure. A good advertisement can trust people to connect ideas. The question is whether the connection leads toward the actual offer.
If leadership cannot agree on that offer before the research starts, the assignment belongs partly in brand strategy and positioning. Asking viewers to settle an internal product disagreement will produce an impressive amount of feedback about a question the company has not decided how to answer.
- Enjoyment
- Did the person like the creative? Their answer may concern humor, casting, music, visual taste or personal relevance.
- Comprehension
- What did they think the company was offering? Did they identify the right category, audience and material conditions?
- Recall
- What could they retrieve after the ad was removed and, where relevant, after a defined delay? Remembering a slogan does not establish that they interpreted the offer correctly.
- Preference
- Did they want the offer they understood? A clear rejection may be useful evidence of poor fit rather than poor communication.
Please do not solve this by making every ad dull.
A surprising image, an unfinished story or a joke can be exactly right for the assignment. A teaser may deliberately leave a question open. A campaign for familiar customers can rely on context a first introduction cannot. Different pieces have different jobs.
I would judge the creative against its intended role. A launch teaser might need people to recognize the brand and anticipate a reveal. A direct inquiry ad needs enough clarity for the next action to make sense. Neither should inherit the other’s requirements by accident.
For Tuesday, “get your Tuesday back” may still be a useful line. The weakness is asking it to identify the service, describe the outcome and explain the next step by itself. A short category description could carry some of that work without draining the personality from the ad.
The distinction matters particularly in specialist markets. A technical abbreviation can be efficient for an engineer and meaningless to a financial approver. An unfamiliar term can also be part of what the campaign needs to teach. I would ask what this audience already knows, what the ad must establish now and what can reasonably wait until the next interaction.
There is no virtue in explaining every detail at once. There is considerable value in identifying the detail whose absence changes what someone believes they are being offered.
Decide what would count as a consequential misunderstanding.
I would start with the actual commercial facts. For Tuesday, those facts include the type of service, the person it is intended to help, what the engagement produces and what the initial conversation does. I would also name what the advertisement must not imply.
A useful statement might be: “An owner should understand that this is an advisory engagement about administrative work, not software that automatically runs the business.” That gives the review a specific question. “People should find us innovative” does not resolve it.
Some omissions are harmless at this stage. Someone may understand the service without knowing how many interviews are included. Other omissions alter the transaction. Believing a paid assessment is a free software trial is a different problem from not remembering a feature name.
Agree on those distinctions before examining responses. Include the person who knows what delivery can actually provide. A creative team should not have to guess whether a tempting interpretation is a promise the company is willing to honor.
- What is being offered?
- An advisory engagement that examines recurring administrative work and recommends changes.
- Who needs to recognize themselves?
- An owner dealing with repeated coordination and administrative burdens.
- What happens after the response?
- A conversation to assess fit, not immediate access to a tool or a staffed service.
- What would materially mislead?
- Expecting automatic operation, outsourced staff or a free product that replaces the paid engagement.
This working note is intentionally plainer than the advertisement. Its job is to give the team a stable reference when the creative produces several interpretations.
“You understood that, right?” has already supplied an answer.
Choose people whose circumstances resemble the audience the campaign is meant to reach. Separate existing customers from unfamiliar prospects when prior knowledge could change the interpretation. Colleagues can help rehearse the session, but their familiarity is precisely what makes them a poor substitute for an outside audience.
GOV.UK’s interview guidance recommends relevant participants, open and neutral questions, and follow-ups that clarify what someone actually said. Pew Research Center explains why wording, response options and question order can affect answers. Those principles inform the prompts below; neither source validates this fictional campaign.1, 2
“What do you think this company offers?”
Let the person name the category. Do not start by telling them it is a consulting business.“Who do you think this is for?”
Listen for the circumstances or role they identify. A broad answer may deserve a follow-up rather than an immediate failure label.“What would you expect to happen if you responded?”
This can reveal whether the person expects a purchase, trial, download, appointment or something the advertiser never intended.“What in the ad gave you that impression?”
Return to the specific word, image, visual hierarchy or surrounding context. Avoid arguing for the creative.These are suggested research prompts, not a standardized or validated measurement instrument. Adapt them to the question and audience without quietly coaching the desired response.
Test the thing people will encounter.
A headline in a slide deck is not necessarily the advertisement. Account names, supporting captions, subtitles, image crops and button labels can change its meaning. A mobile placement can hide a qualification that is obvious on a large presentation screen.
For a comprehension conversation, I would show the intended placement and record the viewing conditions. If the question concerns sound-off video, test the captions and visual sequence that remain without audio. If a carousel needs a later panel to make sense, observe whether that panel is reached before assuming it completes the explanation.
Accessibility belongs in the same discussion. Information conveyed only by color, tiny type or a spoken qualification may not be available to every intended viewer. Include relevant access needs in the research and make the test materials usable. Do not mistake an inaccessible presentation for a problem with the participant’s understanding.
Comprehension and memory also need different treatment. Asking someone to inspect an ad while answering a question tells you about interpretation with the material available. Removing it and asking what they remember introduces retrieval. A later follow-up introduces a delay. State which one you are examining rather than attaching the word “recall” to all three.
I would not use a universal five-second rule for every assignment. An outdoor sign, a technical demonstration and a long-form creator video invite different encounters. The exposure should answer the research question. If comparing versions, keep the relevant conditions consistent and record exceptions.
Seeing an explanation once can also teach a participant how to read the next version. Sequential comparison can still be useful for discussing alternatives, but it is not an independent first impression of both. For a consequential comparative claim, arrange a suitable research design rather than treating an informal before-and-after conversation as a controlled result.
Keep the answer separate from your interpretation of it.
A useful research note preserves the participant’s words, the material they saw and the follow-up that produced the explanation. Then it records the team’s interpretation separately. This makes it possible to disagree about the conclusion without rewriting the evidence.
The table below shows how I would reason about possible responses to Tuesday. Every response is constructed for this article. None came from a research participant.
| Possible response | What it may indicate | What still needs checking |
|---|---|---|
| “It is an app that does the admin.” | A category mismatch: software rather than an advisory engagement. | Did “fewer tabs,” the visual treatment or prior experience with similar ads supply that belief? |
| “They look at how the business works and tell you what to change.” | A reading broadly consistent with the intended service. | What does the person expect the company to implement, and what do they think the first conversation involves? |
| “I am not sure what they do, but I like the line.” | Enjoyment with unresolved comprehension. | Is uncertainty intentional for this placement, or does it prevent the intended next action? |
| “It is consulting. We already have someone doing that.” | Possible accurate understanding with low relevance. | Does their description match this service, or a different kind of consulting they have assumed? |
Do not reduce all ambiguity to a red mark. “I don’t know” and a confidently wrong interpretation are different observations. So are misunderstanding the category and forgetting the advertiser’s name. The repair depends on the distinction.
A small qualitative round can expose a problem worth fixing or a question worth investigating. It cannot, by itself, tell you how common that interpretation is across the market. Report the participants, recruitment, exposure and observed responses. Avoid dressing an exploratory count as a population estimate.
If the investment requires a reliable estimate or a comparison between audiences, involve research expertise in the sample and study design. The consequences of getting the answer wrong should influence how much evidence you require.
Give the idea enough context to do its job.
Here are two possible revisions of the Tuesday ad. Both keep the benefit, identify the service and make the next step more concrete. They make different creative choices, and neither has been tested.
The first protects the original line and adds the missing category. The second leads with the specific work being offered. A team could investigate whether its intended audience understands either version more accurately without assuming that the more literal version must perform better.
Get your
Tuesday back.
A practical review of the admin work consuming your week. Know what to change, delegate or automate.
Which admin work
still needs you?
We map the recurring tasks, examine the handoffs and recommend where your business can work differently.
Both ads are illustrative. The displayed buttons are examples, not active offers. These revisions are hypotheses about clearer communication, not proven winners.
Even these revisions leave legitimate questions. Does the audience understand what an admin-work review produces? Does “automate” make implementation sound included? Does the button accurately describe the first conversation? Those are useful things to investigate, rather than reasons to declare the exercise finished.
If the company wants to advertise a fixed scope or price, the commercial facts need to be settled and represented accurately. If the service varies by situation, the advertisement should not manufacture a tidy package simply because the layout has room for one.
My preference would be to change the element implicated by the evidence, then examine the resulting interpretation. If the image implies software, swapping two adjectives may accomplish very little. If the category is understood but the next action is mistaken for a purchase, the button and its immediate context deserve attention.
Changing the headline, audience, offer, image and destination together might improve the campaign. It also makes a specific explanation of what caused the improvement much harder. Be clear whether you are repairing a practical problem or trying to learn which element produced an effect.
A useful review ends with a decision somebody can explain.
If appropriate prospects repeatedly identify the wrong category, I would address that before scaling distribution. If they understand the offer but do not want it, I would investigate fit, relevance or the offer itself. Rewriting for “clarity” would be an incomplete response to a clear rejection.
If they recognize the offer but cannot connect it to the advertiser, the creative may need a stronger brand connection. If they cannot make out the qualification at the size shown, the layout needs attention. If the team disagrees about whether their interpretation is actually wrong, return to the commercial facts.
Sometimes the sensible decision is to keep the creative. Research should be allowed to support that outcome. A campaign does not become defective because its maker commissioned a review, and a single unexpected answer is not sufficient reason to dismantle an otherwise coherent idea.
I would record the interpretation we intended, what we observed, the limitations of the evidence, the change we chose and what we want the next review to establish. That record gives the creative team something more useful than “make it clearer.”
For campaign planning, distribution and ongoing creative work, this connects to my paid-social consulting. The specialist question in this article remains comprehension. Determining what caused a sale, reconciling platform reports and estimating incremental effects belongs to the separate work of marketing measurement.
The explanation should survive your absence.
The people making an advertisement deserve room to make something interesting. They also deserve a review process that distinguishes a taste disagreement from a real communication problem.
“I don’t like the color” is a preference. “These prospective buyers thought the offer included a service we do not provide” is a different kind of observation. Neither becomes more useful by pretending to be the other.
What I want is an advertisement that can leave the meeting, encounter someone unfamiliar with the company and carry the intended idea far enough for the next decision. It does not have to say everything. It should give that person a fair chance to understand the thing being offered.
The ad has to work after the person who wrote it stops explaining it.
Keep the wit. Keep the surprising image when it earns its place. Keep the dog if the dog is helping. Then ask someone who could plausibly become a customer what they believe they just saw.
Bring the ad.
Tell me what it needs to mean.
I can help examine the offer, the creative and the questions worth putting to prospective customers before your team decides what to change.
Useful starting material: the actual placement, intended audience, offer details and any customer responses you already have.
Discuss the creative with Rob ↗Research notes
The creative examples, interpretations and proposed review approach are original editorial work. They are not a client case study, a validated testing instrument or reported experiment. These primary sources inform the interviewing and question-design discussion:
- GOV.UK Service Manual: Using in-depth interviews. Relevant participant selection, neutral questioning and follow-up practice.
- Pew Research Center: Writing Survey Questions. How wording, response formats and question order can influence answers.