No matter how scary it is to accept it, AI (artificial intelligence) isn’t going anywhere. If anything, it’s become part of our daily work and will definitely envelop our everyday life.
The market research space, more than any other industry, needs to embrace the latest technological advancements ahead of time if it wants to keep up with consumer trends.
As an innovation-driven company, we’ve been fascinated with AI for quite awhile. After experimenting with different AI-powered tools and leveraging AI in our internal projects, we can confidently say that there is a way to ethically use AI to empower, not replace, market researchers.
AI isn’t about replacing people. It’s about augmenting them. By understanding and identifying tasks that take a lot of time or that people struggle with, we can leverage AI to help our in-house team of insights experts focus on what matters - providing high-quality, accurate, and data driven insights.
It all starts with the right team.
Advanced Analytics and the taming of AI
Dig’s Advanced Analytics team is like that group of crazy-smart scientists you see in the movies. The main purpose of the team is to develop new methods and tools to improve and streamline market research processes.
As Kevin Hare, our EVP, said:
Our Advanced Analytics team took something that usually took us weeks to set up, looked under the hood and built something from the ground up to solve the problem.
The end benefit is that research doesn’t take as much people power and time as it did before. The company then is able to process projects faster and take on more work.
One of the biggest things Advanced Analytics is working on is incorporating AI into our proprietary tools. But what exactly does that actually entail?
Idea Maps via word embeddings
AI can extract the “embeddings” vector of each idea, which are mathematical representations of text, and build an idea map of words across studies.
These kinds of idea maps help us determine similarity, or correlation, between ideas or statements. For example, if you want to understand how a new product performs alongside your existing portfolio, an idea map can show clusters of similarity or reveal how incremental it will be to your product mix.
The Advanced Analytics team is currently exploring further applications of embeddings that would use the similarity of ideas in past research as a way of helping validate early-stage concepts - even before testing. The development of word embeddings opens a whole world of possibilities and efficiencies for researchers.
Open-end text summarization and themes
Dig also uses our own proprietary AI tool to summarize qualitative consumer feedback from open-ended questions. It requires no manual work - the data is sent directly to AI, and it identifies relevant topics, extracts main themes, and summarizes the key ideas it finds in the data. Once that’s all done, you can ask questions in natural language about the data and the results that have been generated, and they’ll be answered in real-time by the AI.
The best thing is that our open-end summarization AI tool doesn’t use ChatGPT’s system. We make sure that our client’s data is secure and protected and that our use of AI is ethical and safe.
Read more about our use of AI and privacy.
Chatbot AI
Another qualitative use of AI is chatbots. We leverage those to help us moderate conversations with participants and ask follow-up questions.
For example, the AI chatbot can prompt respondents to go deeper and elaborate on why they said they didn’t like a product or offer. This helps us avoid the problems of receiving blank, vague, or confusing responses to open-ended questions.
We’ve already tested our AI chatbot in some real qualitative interviews and found that it doesn’t negatively affect the respondent experience. While this AI technology can definitely save time for our qualitative consultants, it is their expertise that guides how the chatbot will interact with participants.
Read more: The Future of Qual
Upsiide AI generator
After spending so much time playing with AI for our qualitative work, it was only a matter of time before we developed AI capabilities within our own innovation platform, Upsiide.
We pride Upsiide for being a unique restech platform, built to empower its users to discover and demystify consumer insights. With the introduction of our brand-new AI Idea Generation Tool, users can now ideate, optimize, and test innovative ideas all in one place - how much more agile can you get?
Most importantly, our AI tool isn’t ChatGPT, so there’s no need to worry about data being used for model training without permission. We utilize a more secure version with our AI Prompt Engineers building out the back end for us to get more out of AI, while keeping private data private.
Commentary Writer
Idea generation is just the start of what we’ll be able to do with AI on the Upsiide platform. Our AI team is developing tools that will soon be able to generate thoughts and comments on data in real time. Once generated, just like our qualitative data, those comments can be themed and summarized, and included in your AI report generated by Upsiide. This is just another way for us to make sure that, for clients who want to iterate fast, Upsiide is the tool for the job.
Generating smart synthetic data
With all the hype surrounding AI these days, some people have suggested that soon the research part of market research will be obsolete, and we’ll no longer need to ask people about their thoughts, beliefs, and behaviors.
But we disagree. AI won't replace market research anytime soon.
However, in data science, we often create synthetic datasets to test our models or to create realistic structure in visualizations. We are already using AI to help create smarter synthetic datasets. This improves our models and methods as much as possible before we even begin testing them in the real world.
For example, suppose we have a large proposal that we want to build realistic slides for. We’ve already completed this project for a retail client, but our new client is in insurance. We obviously we want to make it relevant for our client’s business, but it is just a proposal so we don’t need to actually design a questionnaire and conduct fieldwork. Using AI in this way means our client will better understand the value of our analysis if we can show example results.
Chatting with your data
ChatGPT allows us to converse with AI. Bing allows us to do smart AI lookup across the web. But what can we do with our own data? Partitions between sources of information make that information less useful, whether that separation is different files, cloud storage systems, or the brains of the people at a company. But what if you had an assistant who would answer your questions based on all the reports you wrote in 2021, or all the RFPs written over the last ten years?
We will continue to make more and more information accessible to a smarter AI that can provide you with better and more detailed information. All in pursuit of the main goal of market research, to give clients the right answers to the right questions at the right time.
Looking into the future
As we continue to innovate and develop new ways to incorporate AI into market research, we can't help but be excited about what the future holds. With the power of AI, our team at Dig can gain deeper insights into consumer behavior and preferences, streamline research processes, and ultimately help drive better business decisions.
But it’s always important to remember that AI can only glean such results with the help of smart humans.
