4 Ways You Can Use ChatGPT for Market Research

Header image by @jupp via Unsplash

When it comes to online quantitative market research, the process of getting from design to execution can sometimes be very manual and time-intensive. From questionnaire design to survey programming to data analysis, there are a lot of manual steps along the way that can eat up a researchers time.

Generative AI tools like ChatGPT can help automate and streamline a lot of the manual steps in market research. In this article, I’ll show you four examples (with prompts) of how you can use ChatGPT to help you work smarter.

Here’s a rundown of the four use cases I’ll cover in this article.

1. Designing a questionnaire

2. Creating localized brand lists

3. Creating attribute lists

4. Analyzing open-end data

If you’d rather watch and listen, you can check out my video tutorial below. Otherwise, read on for the full list of use cases with example prompts.

1. Designing a questionnaire

When designing questionnaires for online surveys, you need to think like both an artist and an engineer. As an artist, you need to think creatively about how to write questions that map to your business objective. On the other hand, as an engineer, you need to consider technical aspects like question types and question mechanics (e.g. question piping, skip logic, etc).

Here’s an example prompt you can use.

design a survey about brand market share

This prompt can help you quickly design a survey questionnaire. You can also swap out the “brand market share” with a different type of survey. For example, if you want to run a survey to test a new product packaging design, you can modify the prompt as follows:

design a survey to help me get feedback on a new product packaging design

2. Creating localized brand lists

what are the top food and beverage brands in the Philippines

With this prompt, you can use ChatGPT to help you automate secondary research to generate brand lists that are fully localized quickly. So in the example above, you can swap out “Philippines” with whichever country or region you want to focus on.

But the one limitation of the prompt above is that the sector, food and beverage, is quite broad. And so, it can be helpful for you to give ChatGPT more specific instructions. Here’s an example.

what are the top non-alcoholic drink brands in the Philippines

Now that we’ve given ChatGPT a more specific sector to look at (i.e. non-alcoholic drink brands) we’re more likely to get specific results.

3. Creating attributes lists

create a list of features that someone would consider when buying a TV

Like prompt 2 above, ChatGPT is being used here to automate any secondary/desktop research that might be required to source a list of attributes (e.g. product features). This is particularly helpful for sectors that you may not be as familiar with.

4. Analyzing open end data

can you help categorize the following list of open-end data?

In the example, we use ChatGPT to aid our data analysis by crunching some open-end data. This use-case is particularly useful because the standard for qual data analysis is something known as qualitative data coding. Qualitative data coding generally involves reading through a dataset, such as a transcript, and identifying recurring themes (i.e. codes). Manual data coding is a highly manual and time-intensive process, and although ChatGPT isn’t perfect, the results are remarkably good.

That’s it for now. I hope you found this article helpful, and if you haven’t given my video a watch yet, then go check it out!