Editor’s Note: Artificial Intelligence (AI) and its variants, have become incredibly important topics in the world of Insights. You will find that in GRIT, this blog, and in others. There are differing views about what AI represents for Insights, from the fearful to the enthusiastic. Here, Anil Kaul presents an optimistic perspective,
The market research scene hasn’t changed much in the last decade – or even in the last two or three decades. While we’re using technology to make things faster and more efficient, we’re still mostly relying on methods and tools that devour time, money, and employee hours. Yes, there’s more data, and it’s arriving quicker than before. But there’s still a gap between collection and processing. Yes, new consumption tools (like BI dashboards) are emerging. But old-school stalwarts like slide decks and presentations still get the most screen time.
And there’s another problem inherent in our current market research techniques: migrating skill sets. These happen when seasoned MR personnel leave their jobs – a huge amount of specialized knowledge goes with them. That knowledge and skills are hard to replace.
It’s time for a change in how we perform market research. Most people in MR sense that AI is going to be the agent of this change. That raises the question Will AI end market research as we know it? Or will it lead us to a newer, better way?
Can Market Research Be Scalable, Reusable, and Adaptable?
It’s true that some companies are using AI in market research; AI subdisciplines like machine learning and natural language processing have been deployed with remarkable success. But we haven’t seen AI become fully adopted in the market research context. For example, AI could be used to:
- Apply research from one project to other projects.
- Develop “smart” surveys that proactively tailor questions to each participant’s answers.
- Deliver insights within seconds of data collection.
- Combine research data with other sources (IoT, sales, behavioral, etc.) and use it to drive strategic decision making.
- Enable business users to ask questions (in normal, spoken language) and get data-backed answers.
We’re not far from these scenarios becoming a daily reality; as we’ll discuss later in this article, many are already in use. We’re steadily moving toward market research methodologies that are reusable, repeatable, and ready to scale across the enterprise.
So, we can agree that AI has the demonstrated potential to change market research. What’s the future looking like for market researchers? Will AI eat all the MR jobs? No, but it will prompt a change in how jobs are done, particularly as we’re seeing algorithms take over MR’s more repetitive manual tasks.
Algorithms: The New MR Skill
Today’s algorithms can do some of the things that market research team members do, including design surveys and questionnaires and analyze the incoming data. Even more importantly, we can now put business context data into these algorithms, which gives more realistic and accurate results. In time, we can expect to see algorithms that are currently in the development process exceed what human researchers can do in these areas.
This will enhance market research, making it less expensive, less time-consuming, and more widely available. Companies will be able to build or even rent algorithms to do specific tasks; since these algorithms are self-learning and flexible, they can be applied to several different tasks fairly easily.
Where does this leave the human researcher? Instead of laboring away at repetitive tasks, researchers will be able to use their creativity and imagination, focusing on finding and telling stories hidden in the data, adding value to the basic business questions handled by AI, and leveraging what AI doesn’t have – the human brain – to find insights greater than those that AI produces.
In short, we see the future in terms of a human-AI partnership, with each one playing to their own strengths. As a result, market research teams will be able to provide greater value throughout the enterprise. And smaller companies that don’t need or can’t afford a full-time MR team will have access to superior MR capabilities by way of rented algorithms. In fact, we are already seeing this partnership in its early stages.
Real-World Examples of AI-Enhanced Market Research
Monitoring KPIs and strategic decision-making have been the areas most associated with market research, but when we add in AI, we can see other possibilities. Here are a few examples of what’s actually happened when AI is brought into market research:
Uncovering Insights in a Mountain of MR
For one Middle Eastern dairy conglomerate, extracting insights from a mountain of disorganized data was the key challenge. In addition to managing their vast amount of information, they wanted to create a way to make research results easily accessible – essentially building their own internal ‘Google’ page for research studies. Ideally, users would be able to ask a question and get an instant, data-backed answer.
Years of research expertise and deep learning algorithm experience went into the solution, which made both structured and unstructured data into relevant consumable nuggets. First, years of market research data were carefully compiled, cataloged, and curated. An AI platform now organizes and analyzes all the company’s data and makes it deliverable to non-technical users.
Meanwhile, a chatbot powered by natural language processing was trained with different data sources and made available to the employees. This enables any employee to ask a market-research-related question and get an answer in simple, natural language. This tool has effectively extended the reach of MR information in the company – and made accessing hidden insights much quicker and easier.
Combining Internal and External Data for 500,000+ Outlets
A Fortune 100 beverage company needed to identify which of their foodservice accounts, regions, and outlets had the highest potential sales gap for any given category, brand, or SKU. They also needed to recommend the ideal product mix for each one of their 500,000+ outlets.
AI was used to process billions of data points from ten data sources (shipment, sales, consumption, survey, demographic, secondary sources, etc.). To create a fully rounded picture, monthly consumer research data (occasions, consumption habits, competition in the area, etc.) and trade area characteristics (demographics, geographic composition, etc.) were added to the internal data. All of this was fed into the AI-powered recommendation engine, and the results were delivered via dashboard directly to the business users. Thanks to greater insight and speed, the company has realized an estimated 3% in incremental revenue.
AI and MR: A Future Partnership That’s Happening Now
The future of market research is going to be focused on becoming faster, more efficient, and more scalable. It will be a partnership between AI processing and human perception, with algorithms taking over routine tasks and researchers applying their experience to selecting and incorporating the right data and getting specific recommendations from broad insights.
AI is going to propel market research forward into an era of change and expansion. Being able to get data-backed results quicker and more economically than ever before will broaden MR’s in-organization scope and its inter-organization reach. Now’s the time to prepare for and start adapting to this change; market research teams who do will have a significant advantage over those teams who are just trying to catch up to the pace of technology.