One of the biggest challenges to a successful segmentation is getting stakeholders to actually use it. In our experience, stakeholders won’t embrace segments if they can’t relate to them as people. If stakeholders can internalize a segmentation, we see:
- More focused product development – product designers can solve the needs of specific consumers vs. a generic customer
- Better UX research – personas are based on data vs. hypotheses
- Better marketing ROI – messaging and media can be better targeted
- Greater alignment in corporate strategy – different departments can rally around the same priority segments and solve their needs
Our gold standard is to conduct extensive qualitative research to accomplish this, where we can hear directly from consumers who fit the segment and craft immersive videos of their lives. But what if you don’t have the time or money to do additional research? Can AI be a bridge to bringing your segments to life?
With recent advances in AI image generation and AI video, we believe it can.
AI IMAGE GENERATION: READY TO BRING SEGMENTS TO LIFE
Generative AI image generation has improved considerably over the past year and is now surprisingly good. For example, below are the results for the same prompt for a young Japanese woman over the past three years:
While most designers in the insights industry are still using stock photography to pull images of segment personas, we anticipate that this exercise will move to AI design tools in the next year, and we’ve already started that migration. AI images are remarkably realistic, as you can see from the example health and wellness segment persona image below:

Some benefits of using AI to generate segment persona images:
- Create mixed-race personas – mixed race subjects are tough to find in stock photos, but AI has infinite possibilities. The persona above is mixed white/Asian because the segment is majority white, but over indexes Asian.
- Alter the physical characteristics of personas – in the image above, we had AI give him a gut, to better reflect his actual weight
- Put personas in more category-specific situations – to make them more relatable (e.g. shopping, exercising, relaxing at home, etc.), or to create images more relevant to the category. For example, one of our MedTech clients conducted a diabetes-related segmentation and wanted to show a key segment finger pricking to test their glucose levels. This would be impossible with stock photography, but easily done with AI.
AI VIDEO: FAST-IMPROVING METHOD TO BRING SEGMENTS TO LIFE
Video is more immersive than images alone, and gives stakeholders an opportunity to hear from the segment. Right now, our gold standard is to create videos from in-home ethnography footage from qualitative research. Can AI serve the same purpose?
AI video has also made significant progress but is in an earlier stage of development. It’s still impressive. AI video startup Runway recently drew press for inking a deal with movie studios, and OpenAI just released their video creation tool Sora, which is already drawing rave reviews.
We’ve had a lot of success experimenting with these new platforms. For example, we used AI to combine the 2D segment persona image above with a video of one of us to create a video introducing the segment persona, as you can see in the video below.
AI video is progressing rapidly, with new capabilities launching every month. For example, the feature that enabled us to create the video above just launched a few weeks ago!
IMPORTANT CONSIDERATIONS WHEN USING AI TO BRING SEGMENTS TO LIFE
Even with these new AI tools, we still recommend that clients conduct qualitative research after a quantitative segmentation to help bring segments to life. When we need to use AI alone, we’ve learned some key considerations to keep in mind:
- Visualizing segments with AI requires a lot of human analysis – we can’t just load data into AI and have it create great results; we need to summarize what makes the segment unique and differentiated.
- Having physical traits in the segmentation questionnaire is essential – age, gender, and ethnicity are great starting points. Because the segmentation referenced involved diabetes, we included BMI, which was helpful in differentiating segments. Hobbies and tech ownership were also helpful in crafting images. We need to include this data in our prompts.
- Ethically, we think it’s important to subtly call out that the images were created with AI – since we’re trying to bring segments to life, our imagery aims for realism. Our AI designers get a thrill every time they hear a client thought the AI image was a participant from our research; ethically, however, we want to be clear that we used AI.

Want to know more?
Join Vital Findings and Dexcom at The Quirks Event in LA! We will share real-life examples of how we used AI to bring new energy and connection to our client’s segmentation studies. Come to our session, How AI can Bring Insights to Life Through Storytelling, and leave with smart storytelling and visualization techniques using AI.
Don’t have a ticket yet? Sign up now with 15% off using code Q15. We hope to see you there!