YouScan Launches Synthetic Audiences to Test Ad Creatives on Real Consumer Discussions
On November 3, 2026, YouScan introduced Synthetic Audiences, a new tool for modeling how consumers will react to ad messages, visual materials, and marketing concepts before a campaign launches. Unlike synthetic audiences that AI creates from a text description of an imaginary buyer, YouScan builds consumer personas from real posts, comments, and reviews collected by its Social Listening platform. Each persona represents a specific group of discussion participants. A marketer can study its needs, attitudes toward products, preferences, and objections, and then test how such an audience is likely to react to a new ad. At the same time, the modeling results remain AI predictions, not actual consumer responses to an experiment that was run.
How YouScan Builds Synthetic Audiences
Synthetic Audiences uses data from existing Topics, the thematic collections of mentions that companies create in YouScan to monitor brands, competitors, products, or market categories. For example, a food manufacturer can create a Topic that collects discussions of a product category, including consumer reviews, comments about competitors, and posts that mention product characteristics. To build an audience, a marketer selects a Topic, a time range, and, if needed, additional filters. Filtering lets you narrow the source data by language, source, category, and other available parameters. YouScan analyzes the collected discussions and groups their participants by similar behavior patterns, preferences, and interests. A synthetic persona with a description of its characteristics is created for each group. As a result, the company gets not one universal model of the target buyer but a set of consumer segments that reflect the variety of discussions on the chosen topic.
Behavioral Personas: Audience Share, Interests, and Real Mentions
Each synthetic persona contains several blocks of information.
- Consumer behavior description. The system identifies typical interests, preferences, expectations, and negative factors.
- Audience share. YouScan shows what part of the analyzed dataset a given persona represents, along with the number of corresponding mentions.
- Sources and sentiment. For each audience, you can see how discussions are distributed across platforms and the prevailing attitude toward the topic.
- Content preferences. The system analyzes which content formats are typical for members of the group.
- Supporting mentions. Persona descriptions can be matched with the real posts and comments used to build them.
The last capability is especially important. An ordinary AI persona can say that a certain group of buyers prefers discounts, but it gives no verifiable basis for that conclusion. In Synthetic Audiences, a marketer can open the source mentions and check whether consumers really discuss price, promotions, product quality, or other characteristics. However, a persona's share of the analyzed discussions is not equal to its share among all buyers in the market. Active authors of reviews and comments may be represented far more strongly than people who rarely post online.
YouScan Example: Analyzing Pizza Buyers
In its demonstration of the new feature, YouScan used discussions of pizza with spicy honey, Hot Honey Pizza. The company built a Topic that included mentions of the product, discounts, promo codes, and related consumer themes. From the collected data, Synthetic Audiences identified 20 personas. Among the largest groups:
- Pizza Deal Reviewers: 20.3% of the audience, 2,169 mentions.
- Burned Pizza Delivery Complainers: 13.1%, 1,399 mentions.
- Hot Honey Pizza Regulars: 9.6%, 1,023 mentions.
- Spend-the-Day Lifestyle Vloggers: 8%, 853 mentions.
- Domino's Deal Hunters: 5.8%, 622 mentions.
About 42% of discussions came from Google Maps reviews. Other sources included Reddit, YouTube, and TikTok. Positive sentiment appeared in 46% of mentions and negative sentiment in 28%. The analysis showed differences between consumers who value product quality and service, buyers who actively look for good deals, and customers who are unhappy with delivery. For example, the Pizza Deal Reviewers persona is characterized by an interest in tasty food, fast service, and good offers. Its negative factors include high prices without discounts, slow delivery, and poor service. These characteristics were formed from the collected discussions, not from a target audience description set in advance by a marketer.
Creative Simulation: Testing an Ad Message Before Publication
After the audience is built, a marketer can create a Simulation, which models the reaction to a prepared ad. To do this, you select one or more synthetic audiences and upload the materials the company wants to test. These can be ad copy, an image, a new product concept, or a marketing offer. You can also specify the intended publishing channel. YouScan models the reactions of individual personas and produces an overall analysis result. The report shows likely positive and negative reactions, comments on the content, and recommendations for improving the message. For example, one group may rate the price positively but find the offer insufficiently clear. Another may become interested in the product but doubt the claimed benefits.
The system links such reactions to the source discussions so that a marketer can check the context behind the conclusions. This makes it possible to spot potential problems with an ad before a paid campaign launches.
Test Example: The Hot Honey Pizza Offer
In the demonstration scenario, YouScan prepared an Instagram post about a three-week launch of Hot Honey pizza. The campaign included a 20% discount for orders placed through the mobile app. The ad copy and an image were used for the simulation. YouScan deliberately left a specific promo code out of both the text and the visual. During the check, Synthetic Audiences found this gap: buyers were promised a discount with a promo code but were not told where to get it or how to use it. In addition, the system suggested the following:
- Put the promo code directly in the post.
- Explain the discount terms more clearly.
- Revisit the wording of the offer limited to three weeks.
- Support claims about product quality with specific evidence.
- Check how clearly the ingredient features are described.
In this way, the test revealed not only the overall expected reaction to the creative but also a specific weakness in the ad offer. This is a demonstration of the tool's capabilities, not an experiment with real buyers and not confirmation of conversion growth after applying the recommendations.
Competitor Analysis and Finding New Consumer Segments
Synthetic Audiences can be used for more than testing your own ad materials. Because the source data comes from YouScan Topics, a company can build audiences from discussions of competitors or a whole product category. For example, a brand can compare the characteristics of people who discuss its products with groups of consumers who actively engage with competitors. This helps reveal differences in consumer expectations. One group may choose a product mainly on price, another on quality, and a third on delivery convenience. Such data can be used to develop positioning and to form hypotheses about new target segments. However, Synthetic Audiences does not automatically determine the commercial potential of each segment.
The number of discussions and the share of mentions do not allow you to directly calculate market size, purchasing power, or a sales forecast. Such estimates require additional data on purchasing behavior and commercial results.
Testing Marketing Hypotheses Against Real Discussions
Another use of Synthetic Audiences is to compare a marketing team's assumptions with what consumers discuss without a researcher involved. For example, survey results may show that buyers choose a certain product solely because of its low price. However, an analysis of reviews and social posts may reveal that quality, ease of use, and speed of service are discussed as often as cost. Based on this data, a marketer can refine the hypothesis and test several positioning options. Synthetic Audiences also helps uncover less obvious groups of consumers that were not included in the original research plan. At the same time, spontaneous online discussions have their own limits: they more often represent people with strong positive or negative impressions, not necessarily typical buyers. That is why it makes sense to use YouScan results as an additional source of Consumer Insights, not as a full replacement for quantitative research, interviews, and testing ad campaigns on a real audience.
Which Capabilities Are Still in Development
In its November 3 announcement, YouScan outlined the further development of Synthetic Audiences. The company plans to add:
- Chat: a conversational interface for interacting with synthetic personas.
- Comparative Analysis: comparing how different audiences react to the same material.
- Export: exporting audiences and the results of working with them.
These capabilities are presented as future updates, not as features already available at the initial launch. Synthetic Audiences was released as an add-on module for YouScan. At the time of the announcement, the company did not disclose public pricing or exact connection terms. New users can request a product demo.
Using Synthetic Audiences in E-commerce, Customer Services, and B2B
E-commerce. Online stores and consumer brands can use the tool when preparing product launches, seasonal promotions, and ad campaigns. For example, before launching a new collection, you can check which product features generate interest, which discount wordings consumers understand, and which objections come up when they first see the offer. Customer Services. Service companies can analyze typical complaints, expectations, and reasons for choosing a service provider. Synthetic Audiences lets you check how convincingly an ad message answers questions about price, service quality, timing, and service terms. B2B. For companies with a long sales cycle, the tool can be useful when researching public discussions of software, technology, and professional services.
However, the ability to model the reaction of a complex Buying Committee is limited by the volume and quality of available discussions. A synthetic persona does not replace interviews with specific participants in a corporate buying process.
YouScan Combines Social Listening, Visual Analytics, and Consumer Insights
YouScan is a social media monitoring and consumer discussion analysis platform for studying brand perception, market trends, and the competitive environment. The system collects public mentions from social platforms, forums, review sites, and other online sources. Marketers can track discussion trends, assess sentiment, analyze the reasons behind negative reactions, and spot changes in consumer preferences. One of YouScan's distinctive features is Visual Insights, an image analysis technology that detects brand logos, objects, scenes, and other characteristics of visual posts. This helps study user-generated content even when a brand appears in an image but is not mentioned in the text.
The additional product Insights Copilot provides a conversational interface for analyzing large sets of mentions. A user can ask questions about why attitudes toward a brand changed, compare competitors, and get analytical conclusions with the ability to check the source posts. Synthetic Audiences extends this infrastructure with a new way to use Social Listening Data. While YouScan's traditional analytics explains what consumers have already said about a brand or category, the new tool lets you model their possible reaction to material that has not yet been published. For marketing teams, this adds a step for testing hypotheses between audience research and an ad campaign launch.
At the same time, it is important to keep a clear distinction: the source posts represent real statements by users, while the reactions of synthetic personas are the result of AI modeling. That is why Synthetic Audiences should be treated as a tool for preliminary research and creative refinement, and its recommendations should be confirmed with real results of ad tests, conversions, and commercial metrics.
Trust Us Transform Your Performance Marketing.
Pre-launch testing only helps when the insights reach your ad strategy and your results are measured against real conversions. We build paid social and search campaigns around audience research and set up the tracking that shows what each creative really delivers.
Plan My Campaigns


