Discover how AI synthetic focus groups can transform your market research. We explore building personas, querying AI and refining ideas for compelling content and news stories. Unlock deeper audience insights with these skills.
These tips are part of the session "AI as Strategic Partner and Growth Engine," where our speaker, Laura Macdonald, Chief Growth Officer at Hotwire, helps PR professionals learn the latest AI tactics help with audience and market research.
This session was part of the PRNEWS PRO Online Training Workshop: "The AI Shift: Practical Strategies for PR Leaders."
Watch the full session at this link.
Full transcript:
[LAURA MACDONALD]
One area we've started to find, you know, more interest and, and more work being done is with using AI as well to build synthetic focus groups. how many of us have clients that have a bunch of research into their target audience, whether that's surveys you've done in the past for mediogenic reasons or persona information that your marketing team has built? We can use all of that to build a synthetic focus group.
It's not just as simple as putting that information into like a custom GPT or something. Our AI labs team really does have to do some work to map, you know, all of the kind of context and that information, responses from the surveys and things against a kind of normal distribution of people, the kind of traditional bell curve that we're all familiar with, to identify like clusters of the personas, but also the, the volume of those individual kind of clusters as well. And then from that, you can create, 1, that focus group, but also individual kind of synthetic personas aligned to each cluster.
And then taking this even further from going beyond just the information that we have from our clients or we have from, you know, other research, we can also start to build synthetic focus groups what the AI knows about these people too. I'm not sure how many of you, um, did that when it was going around LinkedIn, but if you ask ChatGPT what it knows about you, it's like eerily accurate from, you know, job that you have, the type of work that you do, personality traits. And so we can use that, um, to really feed into these synthetic focus groups as well.
So let's look at this for a transportation client that really wanted to be seen as a government technology company, not transportation. And that was particularly because they really cared about, you know, boosting their profile, um, when it came to retail investors and thinking that, you know, GovTech, you know, stocks typically trade higher than, um, transportation ones. And so by using our, our tool kind of Hotwire Spark, we were able to use AI agents to act as that specific kind of persona and then query, you know, what questions is this persona most likely to ask a specific LLM and what answers are they receiving.
And so what this allowed us to see is that even though the company is positioning itself as a mission-critical technology platform, ultimately what retail investors really do care about and the questions that they're asking LLMs today is actually how should they value GovTech investments and what are the risks of investing into GovTech.
And we can use these focus groups in lots of different ways. You know, firstly, just a gut check. Would this topic or message be of interest to our target audience? And as you saw with the transportation earlier, maybe a topic that we really care about or our clients really care about just isn't one that our target audience does.
Secondly, will this creative idea resonate with our target audience? How would they likely react? And being able to run ideas by our synthetic focus groups, like we used to be able to do with real-life focus groups, um, can help us weed out those kind of creative ideas that just won't resonate, really allow us to refine our proposal and check our ideas before we present them to clients, or indeed take them live. third one that we've also started using these for is to help us directionally with future research.
If we ask these questions, what would the likely response be, and is that interesting? We know that when we're running those, you know, surveys or polls for kind of, you know, ultimately media relations purposes, we often have limited questions that we can ask, and we want to make sure that the ones we ask get us to that interesting headline, that interesting news story that media will want to write. So being able to see ahead of time what would be the most likely responses, um, which ones are the most interesting, um, and which ones give us something that we already know can be incredibly beneficial when it comes to developing kind of, you know, future news story ideas as well.
Produced by: PRNEWS