Attention must make the brand noticeable

An advertisement can attract interest while leaving viewers unaware of the advertiser behind it. For marketers, that creates a spending risk: people may see the content without noticing its brand or message. Before committing media spending, teams therefore need to assess whether the creative attracts attention, makes the advertiser recognisable and supports the campaign’s intended result.

In a fast-scrolling social feed, attention comes first because viewers can pass an advertisement before engaging with its message. Holding their interest, though, leaves a separate question: does the brand register? A logo can appear on screen without viewers noticing it, so pre-launch evaluation needs to examine both stages to judge whether the attention teams expect to buy gives the brand a chance to be noticed.

Pre-testing connects attention with the intended response

These separate requirements shape the questions a creative review should ask. Speaking at DMWF North America, Karel Ellis-Gray, Head of Customer Success at BrainSuite, described the company’s approach: “Does it grab attention? Have I landed my brand? How persuasive is it? Is this asset emotionally engaging? Is there a call to action for me?” Each question addresses a different part of the viewer’s response, so a strong assessment on attention leaves the others open.

BrainSuite has a commercial interest in marketers adopting this approach because it offers BrainSuite AI, a neuroscience-based predictive AI product for assessing advertising assets before launch. Neuroscience studies the nervous system, including how bodily responses relate to information processing and decisions; predictive AI uses computational methods to anticipate responses. Ellis-Gray presents these methods as evidence teams can add to their creative judgment.

In the workflow Ellis-Gray describes, marketers submit assets before committing campaign spending, then use the assessments to identify changes. Attention concerns whether the content attracts interest, while brand registration concerns whether viewers notice the advertiser. Persuasion addresses influence on the viewer. Emotional engagement examines the feelings the creative evokes, and a call to action tells viewers what to do next.

Ellis-Gray puts attention first, followed by brand registration and consideration of persuasion, without treating every response as a compulsory sequence. Emotional engagement and the call to action extend the review to what an attentive viewer might feel and do. Teams can use these questions to examine an asset while changes remain possible.

For social teams, Ellis-Gray proposes building that review into the publishing workflow: “through the power of predictive AI and pre-testing, you are able to integrate both into a channel such as social media, basically pre-testing all of your assets before they go live.” In his account, testing at this stage lets teams address weaknesses in attention or branding before paying for distribution. With creator content, when the brand appears becomes a particularly important decision.

Okoone experts
LET'S TALK!

A project in mind?
Schedule a 30-minute meeting with us.

Senior experts helping you move faster across product, engineering, cloud & AI.

Please enter a valid business email address.

Creator content makes brand timing a practical test

Brand timing can be difficult in creator content because an unobtrusive commercial presence can be part of its appeal. Ellis-Gray points to authenticity and the creator’s relationship with their audience as reasons teams keep branding subtle. In some videos, the advertiser appears several minutes after the content begins.

During that delay, the advertiser depends on viewers continuing to watch. People who leave earlier may have engaged with the creator without encountering the brand, so timing affects the opportunity for recognition. As Ellis-Gray put it: “Despite the channel, whether that is a branded content piece or the new influx of creator content that we have, it’s really important that you’re still landing the brand.”

To assess that opportunity, Ellis-Gray says BrainSuite examines attention alongside audience drop-off, the points at which viewers stop watching before the content ends. The brand must be noticeable, and viewers must still be there when it appears. A review can therefore consider when the advertiser enters the content and whether that moment falls within the audience’s likely viewing time.

Ellis-Gray describes how users have applied that analysis: “Learnings that our users have taken are encouraging teams to show that brand a little bit earlier. The kind of thinking behind creator content is this authenticity. They have their audience. It’s not so in your face, if you like.” The creative task is to find an earlier introduction that fits the creator’s approach, since a more immediate commercial presence could weaken the unobtrusive quality the team wants to preserve.

An earlier introduction gives teams a specific change to evaluate before publication and check against results after the campaign. The review then moves from checking whether branding appears to judging whether viewers have a reasonable chance to notice it. But evaluating the benefit of that change requires clarity about what a predictive assessment measures.

The 81% finding measures advertising preference

Advertising preference is one outcome researchers have tested using neuroscience-related information and machine learning. A recent study reported 81% accuracy for a Random Forest model predicting which advertisements participants preferred. Preference differs from brand recognition, purchases and overall campaign success, so the result needs to be interpreted in those terms.

To make that prediction, researchers showed participants a series of videos and recorded their bodily responses. They used Facial Expression Analysis (FEA) to examine facial expressions and Electrodermal Activity (EDA) to measure changes in the skin’s electrical activity. Machine learning, which uses patterns in data to make predictions, then combined this response information.

The Random Forest (RF) model produced its predictions by combining multiple decision trees, branching sets of decision rules. The study also associated joy with engagement in the advertisements tested. These findings show how measured responses can help anticipate advertising preferences, with the interpretation tied to the participants, advertisements and outcome examined.

The study’s setting also limits that interpretation: it covered one sector and used a relatively small sample. The reported accuracy belongs to that preference-prediction experiment. BrainSuite’s brand-registration assessments are a separate commercial application, so the research result should remain distinct from claims about their accuracy or effectiveness.

For a marketing team, the practical question is how the predicted response relates to the campaign’s purpose. Preference can be useful, but the intended result may be a click, a completed view or a purchase. Teams therefore need to identify that result before deciding how much weight to give a pre-test judgment.

Check pre-test judgments against the campaign’s goal

Checking a prediction starts with choosing a performance indicator that represents the campaign’s aim. Click-through rate measures the share of advertising exposures that produce clicks; view-through concerns whether people continue watching; purchases capture a buying action. A pre-test becomes commercially useful when teams can examine how its assessment relates to the outcome they sought.

Ellis-Gray describes that objective-specific approach: “We can say, what is the campaign goal here? Are people looking for a higher click-through rate? Is the important metric for one of our customers view-through? Are they really looking for a purchase? And with some of the studies that we’ve done, with the kind of advantages of having that performance data, you’re then able to look back.” His account includes retrospective analysis in the assessment process, allowing teams to compare an earlier judgment with subsequent performance.

BrainSuite describes combining its pre-test effectiveness scores with data collected after a campaign runs. Teams can compare those scores with the chosen outcome to examine whether creative assessed as engaging and noticeable produced the result they wanted. Checking a specific pre-launch expectation against performance helps teams decide how much weight to give that signal in later work.

Across successive campaigns, BrainSuite says historical performance data can help teams identify patterns associated with stronger outcomes. Teams can use those patterns to review new creative and decide where to invest media spending. As comparisons accumulate, actual campaign results provide a basis for judging which pre-test signals are useful for the business’s objectives.

Beyond those immediate objectives, BrainSuite says historical patterns may reveal factors associated with wider reach, including the possibility of a product going viral. Such associations can inform creative choices and potentially improve media-spending efficiency, while virality and campaign success remain uncertain. Teams can form an expectation about reach from those patterns, then test it against the next campaign’s performance.

Key takeaways for decision-makers

  • Brand recognition needs a separate check: Marketing teams evaluating creative before media spending need to assess whether viewers notice the advertiser as well as the content. Attention creates an opportunity for the brand to register.
  • Pre-testing needs clear criteria: Creative teams using predictive AI can assess attention, brand recognition, persuasion, emotional engagement and the call to action before launch. These assessments help identify changes while the assets remain editable.
  • Creator branding depends on viewing time: Social teams reviewing creator videos need to compare brand placement with audience drop-off. Testing an earlier brand introduction helps assess whether more viewers encounter the advertiser while preserving the creator’s approach.
  • Prediction accuracy has a defined scope: The reported 81% accuracy concerns advertising preference in a small, single-sector study. Buyers evaluating predictive tools need evidence tied to the outcome they want to predict and the application they plan to use.
  • Campaign results provide the performance check: Marketing analytics teams can compare pre-test scores with the campaign’s chosen outcome, such as clicks, continued viewing or purchases. Repeated comparisons help establish which signals deserve weight in future creative and spending decisions.

Alexander Procter

November 16, 2026

8 Min

Okoone experts
LET'S TALK!

A project in mind?
Schedule a 30-minute meeting with us.

Senior experts helping you move faster across product, engineering, cloud & AI.

Please enter a valid business email address.