How Synthetic Personas Help Shape Better Video Marketing

Audience personas have been around for years.

Most of them are useful in theory, then disappear into a strategy deck and never get used again.

Synthetic personas are different when they are used properly. They are AI-assisted audience profiles that help marketers think through how different types of people might respond to a brand, message or content idea.

They are not real people. They are not a shortcut to perfect audience insight. They are definitely not reliable enough to replace research.

But they can be a helpful thinking tool.

When we were working with a wellbeing client, we used synthetic personas to explore how different audiences might engage with their video content. The client is an evidence-led functional music brand, sitting at the intersection of sound, wellbeing, science and culture. That means the audience is not one simple group.

Some people come to functional music because they want to sleep better. Some want to focus. Some are interested in the science. Some are drawn to the cultural and emotional experience of sound.

Trying to speak to all of those people in the same way would make the content too broad. So we created synthetic personas to give each audience type a clearer role in the video strategy.

What are synthetic personas?

Synthetic personas are fictional audience profiles created using a mix of research, strategy, market understanding and AI-assisted modelling.

They are designed to represent different audience types.

A useful synthetic persona might explore:

What the audience cares about
What questions they are asking
What might make them trust a brand
What might make them sceptical
What kind of content they prefer
What they need to see before taking action

The point is not to pretend these personas are real people. The point is to use them to test your thinking.

Instead of asking, “Do we like this idea?”, you can ask, “Would this audience care?”

That is a much better question.

Why synthetic personas are useful for video strategy

Video has a lot of moving parts.

You need to make decisions about the concept, script, structure, visuals, interviews, edit, titles, thumbnails, distribution and measurement. Without clear audience thinking, it is easy to make content that looks good but does not have a sharp enough job to do.

Synthetic personas help by giving each video a clearer purpose.

A video for the Evidence-Led Optimiser might need a clear question and a credible expert.

A video for the Emotional Wellness Seeker might need warmth, simplicity and a more human story.

A video for the Cultural Tastemaker might need a stronger visual world, stronger sound design and a sense of cultural identity.

Why they are useful for marketing more broadly

Synthetic personas are useful because they help marketers make clearer decisions.

A lot of marketing falls apart when the audience is too loosely defined. The messaging becomes generic, the content tries to say too much, and every video ends up targeting “everyone”.

Personas help marketers think about what a particular audience actually needs in order to care, trust the brand or take action. That can affect everything from the hook of a video to the tone of a landing page.

They are also useful because different audiences often need completely different types of content. One audience might respond to educational explainers, while another responds better to emotional storytelling or social proof.

Synthetic personas can also help teams plan content ecosystems more effectively. Instead of expecting one video or campaign to do everything, marketers can create different pieces of content for different audience needs and stages of interest.

That usually leads to sharper messaging, more focused creative decisions and content that feels more relevant to the people it is trying to reach.

How we created synthetic personas

We started with the brand strategy.

Our client sits across sound, wellbeing, science and culture, so the audience needed to be segmented by motivation rather than simple demographics.

From there, we looked at:

• The different reasons someone might engage with functional music
• The types of video content that would serve each audience
• Competitor and category signals
• Likely search behaviour
• Audience objections and trust barriers
• The emotional role of sound in people’s lives

AI then helped us model each audience more clearly.

We used the synthetic personas to pressure-test content ideas. For example:

• Would this video title appeal to them?
• Would this opening hook make them keep watching?
• Would they need more evidence?
• Would this feel too clinical?
• Would this feel too vague?
• What would make them share it?

This made the personas practical. They became a way to judge creative decisions.

How synthetic personas support SEO and LLM optimisation

People are no longer only searching through Google. They are asking AI tools for explanations, recommendations and comparisons. That means content needs to be clear, structured and genuinely useful.

When you understand what different audiences are worried about, curious about or trying to solve, you get a much clearer idea of the questions they are actually searching for.

That is useful for SEO, because good search content usually starts with understanding intent. Someone looking for scientific proof will search very differently from someone looking for emotional support or inspiration.

It is also becoming more important for LLM optimisation. AI tools are increasingly pulling answers from content that is clear, structured and directly relevant to the question being asked.

Synthetic personas help marketers think beyond keywords and focus more on audience intent. That can improve things like article structure, FAQs, video titles, transcripts and supporting content around a campaign.

The risks of synthetic personas

Synthetic personas are useful, but only if you are honest about their limits – they are not reliable evidence.

They are based on assumptions, even when those assumptions are informed by research. AI can make those assumptions sound more certain than they really are., which creates risk. 

A persona might suggest that an audience wants scientific detail, but real viewing data might show that they prefer personal stories. It might suggest that a certain title will work, but YouTube analytics may say otherwise.

Synthetic personas should be treated as hypotheses. They can help you plan, but they cannot tell you what will definitely work.

The right way to use synthetic personas

The best way to use synthetic personas is as a thinking tool, not a source of truth.

They are useful for shaping ideas, testing messaging and spotting gaps in a strategy. They can help marketers pressure-test whether a piece of content actually makes sense for the audience it is trying to reach.

But they should not be treated as fully reliable audience research.

The most effective approach is to use personas early in the planning process, then validate the thinking with real-world data. That could be search behaviour, video retention, comments, conversions, customer interviews or campaign performance.

The danger comes when teams become too confident in the personas themselves. AI can make assumptions sound convincing, even when they are not accurate.

Synthetic personas work best when they stay flexible. They should evolve as marketers learn more about how real audiences actually behave.

The personas we created

One of the most useful parts of synthetic personas is that they help break audiences down by motivation rather than just demographics.

Instead of thinking in broad categories like age or job title, you can start thinking about why someone is engaging with the brand in the first place.

For this project, we focused on three broad audience types that are common across a lot of wellness, lifestyle and education-led brands. We’ve included this in detail under each of these headings, but these are just examples of what we did for a specific client use case. If you would like some advice on what would work best for your organisation, we’re always happy to chat!

Final thoughts

Synthetic personas are not magic. They are not a replacement for research. They are not reliable enough to tell you exactly what your audience will do.

But they are useful.

They help marketing teams think more clearly about audience needs, especially when planning video content. They can improve creative decisions, sharpen SEO strategy, support LLM optimisation and help brands build a more connected content ecosystem.

For the client, they helped turn a broad audience opportunity into a clearer set of content directions.

The most important thing is to use them with the right level of caution.
Synthetic personas are a way to ask better questions before you create content.