How Marketing Teams Can Finally Get Their Own Customer Insights With The Help of AI

Marketing teams’ mission is to speak to their audience in a compelling and effective way. As every great marketer knows, that often comes from deeply understanding your audience. And yet, the vast majority of marketers don’t have the resources for consumer research.

Instead, they are forced to rely on Gartner reports or desk research collected by third parties and approximate an understanding of their user. In a world of constrained resources, it rarely makes sense to invest in traditional, time-consuming, and often very expensive customer research initiatives. But sadly, where does this reality leave marketers? In the dark.

AI is changing this reality

AI in research is finally able to reduce the ‘cost-per-insight’ – the cost it takes to learn deep and relevant insights about your customer. This is because large language models (LLMs) handle the messiness of language and speech, which is the key to deeper, qualitative insights.

This advancement means a marketing team that has limited or even no research-oriented team members can actually get their own customer insights.

Marketing Technology News: MarTech Interview with Elad Stern, President and co-founder @ Odeeo

So how can you get low-cost insights with AI

There are a lot of new tools out there. And while most marketers are likely used to playing with ChatGPT, they may not yet be familiar with how innovative researchers are starting to take advantage of AI to better understand their audiences.

Here are three ways marketers can use AI right now to get consumer insights at scale with the depth of a one-to-one interview.

1. Synthetic testing. It may sound wild, but you can now simulate real user feedback by using AI to stand-in for consumers – often known as synthetic users or personas. This is best for rapid ideation, testing, and iteration, and is by far the fastest way to use AI for insights. For example, Apple probably could have avoided the backlash over its Crush! iPad Pro campaign if it had just been tested with some AI-users.There are lots of technologies popping up to do just this, from startups like Synthetic Users, service firms like Rehab AI, or even Nielsen.
2. Use AI to interview consumers directly. If you want to hear directly from consumers but don’t have the team of researchers to go out and interview them individually, you can just have AI do it! When you think about what ChatGPT is best at, what comes to mind? Conversation. Which, as it turns out, is the basis of qualitative research.AI-powered in-depth interviews are like a survey in that they can scale endlessly without any additional work per participant. But unlike a survey, it digs deep into the participant’s experience, motivations, and environment over video conversations. And for many users they will share even more with AI as they don’t feel like they’re being judged the way a human interviewer might on more personal topics.There are a handful of new companies offering tools like this, including Outset.
3. Use AI to synthesize unstructured data to tell you what’s important.  No matter how you are collecting qualitative data, it is, by definition, unstructured. It’s on us (marketers, researchers, product builders) to sort through all the messy data to create meaning. Ultimately, we want to deeply know our audience – that’s hard to do with reams of transcripts.AI has become quite effective at structuring unstructured data, organizing verbatims, identifying themes, and offering a more thorough picture of your audience. Just about every research tool (including Outset, of course!) is finding creative ways to incorporate AI into analysis – from chat-with-your data to automated themes to broad sentiment analysis. Even ChatGPT can work reasonably well (pro-tip: the more specific and narrow your analysis question, the better the results!).

Conclusion

Lots of researchers are picking up these tools as we speak. And it’s incredibly invigorating to see.

But what’s often overlooked is the broader impact. Reducing the ‘cost-per-insight’ doesn’t just mean the same people do the same research for less money. It means that more people are empowered to go find, discover and act on real customer insights. In particular, marketers.

We’re only now seeing the first wave of impact. But I’m excited to see what the next level of customer-centricity looks like for the world of marketing and how AI can be embraced by marketing teams to get those in-depth insights quickly, and at a fraction of the price of traditional methods. This is how AI can deliver real, tangible results and a strong return on investment in consumer marketing.

Marketing Technology News: Putting Tech to work on your Unique Assets