How Is Generative AI Changing Market Research for CMOs?

By Ron Howard, Chief Executive Officer, Mercury Analytics

There’s so much talk about generative AI. How it is transforming every aspect of our lives, the industries we work in, the jobs we do. In some cases, this is an over-exaggeration.

However, when it comes to research, it isn’t.

For CMOs, traditional approaches to market research are being rewritten. What was once seen as too difficult, too time intensive, too complex to do, is now possible with the proper and careful application of AI.

It is truly transformative, giving CMOs a cutting-edge new weapon to inform marketing strategies in a way that hasn’t been possible before – all in moments, for pennies, with greater flexibility and precision.

Generative AI enables focus group level insight from quantitative research. It gives CMOs more power than ever to understand what their customers think about their products and brands, and why; identify customer preferences within segments; make sense of complex and nuanced issues.

Whether you want to understand how your customers feel about social issues, what comes to mind when they think of your brand, how best to respond to an emerging crisis, how to test messaging before it goes public, now you can.

And you can do so in a way that is extremely cost-effective, rapid, with no bias or analyst fatigue.

In the past, quantitative market research was limited in many ways. Limited by your ability to ask the right questions, limited by your ability to define the potential responses, limited by your ability to code critically insightful open-ended responses, limited by available time and the human and financial resources you can spend on analysis. The list goes on.

However, the value of quantitative research has always been one of the bedrocks of marketing research. With generative AI its value increases significantly.

Enabling deeper and richer insight from open-ended questions

Understanding quantitative data is easy. Analyzing the results by segment is straightforward. This is in contrast to qualitative research where we tend to be dealing with perceptions, emotional responses, and nuance. This is much harder to do.

The most common way to get qualitative insight from quantitative research is through open-end questions. However, researchers tend to shy away from including valuable “why” style questions as they are hard to analyze. To make sense of responses requires a well-trained analyst and many hours or days to code responses into categories, and no two analysts would code responses identically in or wouldn’t become fatigued coding 1,000 or 2,000 responses.

This is no longer the case with generative AI tools. The responses from a sample of 1,000 people can be analyzed in seconds. Whether you are looking for the top ten themes, the most common words used, want an overall or emotional summary of all, or just a segment of survey participants, this can be done by applying AI, ensuring you do not miss the golden nugget of insight.

This is transformative.     

A genuine silver bullet that is here today

The ability to ask more insightful open-ended questions within quantitative surveys enables greater engagement from respondents as people want to express their opinions freely, beyond the constraints of predefined answer choices that they must select from.

Responses are conversational, natural and emotional.

We recently conducted a research study on Unity in America in this election year with over 1,000 voters, to explore voter attitudes to the current state of politics. The study asked if a candidate pledged to communicate their views on issues without insulting members of the other party, negotiate hard but find compromise even when some issues will be disagreeable, and pledge I am an American first, I support my party second, would you be more or less likely to support them. It also introduced a “Unity Pledge” that would commit a political candidate to clear communication, collaboration, and a willingness to compromise with their counterparts across the aisle. Along with answering a series of questions about feelings towards the pledge, respondents used Mercury’s Text Highlighting tool to indicate the specific words or phrases within the candidate’s statement they either agreed or disagreed with.

This enabled us to explore people’s emotional responses and feelings to the statement. Through the application of AI we were able to make sense of these insights, comparing different demographic group responses in seconds.

Likewise, when Black Lives Matter emerged as a movement, we conducted a study where we asked people ‘In what ways do you think there is racial injustice towards the black community in America?’

This complex question received over 1800 open-ended responses. The responses were conversational, natural, and emotional. We recently went back to this study and applied AI. We were able to analyze those responses quickly and easily. Understanding people’s emotional response to the question, the key themes based on different segments, and how each segment compared to another. This simply wasn’t possible when we originally did the survey, without extensive budgets and intensive analyst resources coding each response.

AI can be applied beyond quantitative research

But it isn’t only in the realm of market research where AI is transforming the access to information that CMOs rely on.

All brands, especially consumer brands, have a wealth of information and insight from customers in the form of reviews, social media, feedback forms, forums and more. However, the challenge is often getting to the insights that matter. This is where generative AI can be again transformative. By ingesting, processing, and making sense of vast amounts of unstructured data, it is possible to extract key insights that can provide real value to the CMO and marketing teams, all in record time.

Rather than trying to make sense of what your customers are saying to you due to the overwhelming volume of information, you can finally understand precisely what they are saying and plan well-targeted actions and responses. And you can do it every day, or week, or whenever the need arises.

These are just some ways in which generative AI is changing how CMOs and senior marketers can use research. But it is only the tip of the iceberg. It can be used to better process input from focus groups; to guide what research should focus on from the outset by bringing together what is top of mind for customers; and counter-intuitively, bring a more human touch to research.

And unlike many silver bullets, research can be conducted more cost-effectively, with reduced analyst resources, and in shortened timelines. A sort of holy trinity.

There is no question, that generative AI is a real game-changer when it comes to research.

If you think it is too good to be true. I’d ask you to think again. This is one of those research innovations that really is as good as it claims to be.

About the Author

Ron Howard is Chief Executive Officer of Mercury Analytics, an innovative full-service research firm in Washington, D.C. Ron created Mercury Analytics to change the research landscape using the power of technology to modernize how research is conducted. The firm works with major brands and corporations to conduct quantitative and qualitative research, media testing, applying MercuryAI to get the most from research.


Tags: AI