How to Use AI for Market Research Tasks

11 min. read

Updated November 22, 2024

As I’ve explored how AI can help with business planning, I’ve wondered how useful tools like ChatGPT can be for market research.

While AI can help with writing parts of a business plan, I’ve been skeptical about how helpful it would be for market research.

That’s because market research typically involves gathering data – both from data sources with information about market sizes, and directly from potential customers to determine specific wants and needs.

This kind of factual data gathering has (so far) not been a strong point for AI. Tools like ChatGPT and Google’s Bard tend to “hallucinate” and make up information as they compose their answers.

To state the obvious, you never want fictional market data in your business plan. You wouldn’t want to misrepresent your business opportunity to investors, and you certainly can’t afford to build a business based on inaccurate data.

But with the recent improvements to the various AI tools out there, I thought I’d test AI on a market research task to see how the most popular tools do. I decided to work through the stages of a market research project for a fitness studio using the three most popular AI chatbots (ChatGPT, Bard, and Claude) to compare the results. (Since I’m a paying ChatGPT subscriber, I used ChatGPT 4 for this exercise, along with the publicly available Bard and Claude interfaces.)

Here are the main market research tasks that I’ll be digging into:

1.Identifying target market segments
2.Determining the size of each market segment
3.Identifying competitors
4.Developing ideas for filling market gaps to differentiate from the competition
5.Understanding the industry and identifying potential opportunities and challenges
6.Creating a survey to gather information from prospective customers
7.Developing target market personas
8.Generating marketing campaign ideas

Let’s get started, and see how AI does with these core market research tasks.

Define Target Market Segments

A first step for any market research project is to define the target markets that your ideal customers fall into. As a quick refresher, a target market is a group of potential customers that you plan to sell to. Each different target market that a company sells to is a “segment.”

Target markets are often defined by either demographics or psychographics, or a combination of the two. Demographics are things such as age, location, gender, and income. Psychographics are interests, hobbies, and habits.

For the fitness studio example, the first step is to brainstorm different potential target markets and define the core demographics and psychographics for each segment.

Using the following prompt generated a list of potential market segments, with their demographics and psychographics defined:

Here is Claude’s response to the same prompt:

And Bard’s:

Each of the AI models did a good job at this task, giving me some ideas of target markets, as well as their demographics and psychographics. Bard and ChatGPT formatted their answers the best and were the most readable, but the information from Claude was useful as well.

Interestingly, there wasn’t much differentiation between the three AI tools in terms of the segments that they picked. They each had a few unique suggestions, but also plenty of overlap. This could be because none of these AI tools are particularly specialized – they’re all trained on the same data sets which means that their core knowledge is pretty similar. It’s as if you asked three business consultants who all have the same education and background the same question – you’d likely get pretty similar results back from each of the consultants.

Verdict: AI chatbots are a useful tool to brainstorm market segments and help you come up with ideas. As an entrepreneur, I wouldn’t start a business and target every market that the bots suggested. That’s because it’s never a good idea to try and sell your product or service to “everyone”. Being focused in the early stages of business will help you stay on track and grow faster. The ideas that AI suggested were a good place to start, though, and got me thinking about the specific markets that would be a good fit for my fitness studio.

Get market size information for each segment

The next step is to understand how big each market segment is, so I can get a better understanding of how many potential customers live near the fitness studio in this project. At this stage, I’m going to narrow down the list of segments to three that this business will target:

•Fitness Enthusiasts and Athletes
•Corporate and Busy Professionals
•Active Older Adults

Here is the prompt that I used:

And here are the responses:

Bard:

Claude:

ChatGPT:

On the surface, the answers from Claude and Bard looked useful. Both provided actual data, and both shared links to sources. But I’ve observed, and some embarrassing real-world examples have shown, that AI tools often present convincing answers with authority – only to be outright wrong.

So, I dug in to verify the data that they provided and, as I feared, this task is where the AI tools really fell flat.

Bard’s data was the most detailed and specific. However, it linked to a study that was behind a paywall, so there was no way to easily verify the data that it provided. It also referenced a website that it said was providing 2023 data. The reality is that the website it referenced had 2021 data. Even worse, the numbers that Bard presented were not to be found anywhere on the page it cited. Bard also confidently stated that there are nearly 70,000 gym memberships in our target market, which has a total population of about 200,000. This just doesn’t seem believable.

Claude also provided data and links to sources. However, several of the links it provided linked to pages that no longer existed. I even checked the Internet Archive for these pages and they seem to never have existed! This is a clear case of AI “hallucination.”

ChatGPT was probably the most “honest” of the three AI chatbots. Instead of providing data, it stated that it couldn’t find specific information and instead provided advice and guidance to do my own research.

Verdict: AI chatbots are not yet close to being reliable enough for factual research and data gathering. They may help you understand how you might go about doing your own market research, and they may even provide you with links to potentially useful data sources. But you should gather data on your own if you want to be sure that what you’re presenting is factually correct.

Identify Competition

Part of understanding your potential customer includes getting an understanding of what other products or services they may be using – this is your competition. AI tools might be more useful at this task than citing accurate data sources.

Prompt:

ChatGPT:

Claude:

Bard:

I expected to get better results from AI on this task. Instead, I was surprised by the variety and inaccuracy of the responses I received.

Bard was the clear winner here. It categorized the results, finding both fitness studios dedicated to classes similar to what my studio offers as well as traditional gyms that include fitness classes as part of their offering. It also offered advice on doing my own research at the end of the response.

ChatGPT surprised me with a very basic response. It listed out a number of big-box gyms and did not provide any truly direct competitors. Honestly, I got better results by just doing a Google search.

Claude was the hands-down loser for this task. Its results turned out to be largely invented. None of the URLs it provided worked, and some of the businesses it listed simply don’t exist – Claude invented them!

Keep in mind that I am using a single conversation thread in each of the three chatbots so that they should have context of the prompts that I’ve provided as well as the responses that they have provided. But the results unfortunately were disappointing despite this level of context. The lack of relevant results and the inclusion of non-existent businesses continued to reinforce the notion that the chatbots just couldn’t be trusted.

Verdict: AI can provide some help here, but similar to the market size task above, you need to verify the data provided and do your own research. For my specific business type, doing research on Google would likely yield as good, or better, results than relying on AI.

Get suggestions for filling market gaps and differentiating from competitors

The purpose of market research is twofold. First, you want to understand your customer so that you can build a successful business. Second, you can use your market research to customize your business offering, helping you stand out from the competition and offer something unique in the market.

In general, I’ve found that AI chatbots are most helpful with these types of creative tasks. They can help generate ideas and spur creativity.

Prompt:

ChatGPT:

Bard:

Claude:

ChatGPT and Bard are the winners of this task. They both provided creative, yet (mostly) realistic ideas for the fitness studio business. Neither directly addressed the existing competition, but the ideas are still interesting and worth looking into.

Claude, on the other hand, was less useful. While it offered a few useful ideas, it strayed away from the concept of a fitness studio and generated ideas that were less realistic to implement.

Verdict: This is the type of task where AI chatbots really shine. They’re great at being creative and coming up with a high volume of ideas very quickly. That said, you can’t just cut and paste this content into a business plan. You have to pick and choose which ideas make sense for you, and perhaps follow up with additional prompts to dig deeper on particular ideas. The list of ideas may even spur your own unique thoughts about how you can differentiate from the competition.

Additional ways to use AI for market research

AI chatbots are a mixed bag for market research. They’re useful for creative tasks and brainstorming, but are less than helpful when it comes to getting specific, data driven research accomplished (as of January 2024).

But that doesn’t mean you shouldn’t use AI at all for marketing tasks. In fact, AI can be extremely helpful at generating ideas and getting marketing projects started. Here are a few other marketing tasks that AI can be helpful with as you work on your market research project:

•Understand your industry and identify potential opportunities and challenges. AI can help you get an understanding of how your industry works and help you develop strategies to thrive. This is useful to better understand the business that you’re in and trends that may be developing within your industry.
•Create a survey. Market research isn’t complete without actually talking to your potential customers. You can use AI to get a jump start on a survey that you can send out to potential or existing customers to better understand their needs, current purchase habits, and other information that may be useful for your business.
•Develop personas. A buyer persona is a description of a potential customer. It helps you transform market research data into a useful profile of someone who might purchase your product or service, informing your marketing and product or service design.
•Generate marketing campaign ideas. As your market research helps you define who your target market is and what their needs are, you can leverage that information to generate campaign ideas to reach your audience through advertising and other methods. AI can help generate ideas and even come up with initial marketing copy for your ads.

As generative AI has taken the world by storm over the past 18 months, it’s been tempting to assume that it can handle virtually any business task. And while AI is certainly extremely helpful for some tasks, it struggles at others. Entrepreneurs are still a necessary part of the equation – questioning data and sorting through ideas to determine which are worth investing time and money in.

Stay tuned for more explorations in AI to uncover how these new tools can help us run better, smarter businesses.

Noah Parsons

Before joining Palo Alto Software, Noah Parsons was an early Internet marketing and product expert in the Silicon Valley. He joined Yahoo! in 1996 as one of its first 101 employees and become Producer of the Yahoo! Employment property as part of the Yahoo! Classifieds team before leaving to serve as Director of Production at Epinions.com. He is a graduate of Princeton University. Noah devotes most of his free time to his three young sons. In the winter you'll find him giving them lessons on the ski slopes, and in summer they're usually involved in a variety of outdoor pursuits. Noah is currently the COO at Palo Alto Software, makers of the online business plan app LivePlan.