AI (or, artificial intelligence) is a topic which is becoming increasingly prevalent. Whilst it is becoming more common within the market research industry, I don't believe it can’t totally replace a researcher and how we think and work (thankfully or we’d all be looking for new jobs!); and on top of that I don't think that it should.
Companies these days are generating more data than ever before so, don't get me wrong, anything that makes our lives easier is great - and as Helene Protopapas discussed the fact that AI can help speed up research is advantageous to us all.
However, there are some reasons why I strongly believe that AI isn’t the ‘produce insight’ button we’ve all secretly been wishing for and that you still need a researcher to deliver the insight that clients want and expect.
Questioning the Data
As Harmony Crawford points out, if you just give someone a load of numbers, they’ll drown in them. Whilst they’ll have the data, they won’t have the insight. To get the latter you need a human brain to question the data.
A researcher can ask those questions that AI tools can’t; they can look further into the survey results, ask those pertinent questions to unearth the ‘why’ from the ‘what’. They can give you true insight. The researcher can decide which cross-tabs to run, which filters they need to apply to be able to really understand the story.
For example, a researcher will see the NPS score and rather than taking it at face value; they’ll look to explore the factors that have contributed to it, build the background story and identify how the business can go forwards making improvements.
With AI you could calculate your NPS score quickly and accurately, but there would be little depth of understanding behind it and the business won’t know how they can go about improving the score – they’d have the result but not the insight.
What About Qual?
At the moment, AI isn’t intelligent enough to be able to handle qual.
It can help out with quant projects and run your basic frequencies for sure. But in the case of qualitative research, where you might have tens of thousands of words to read through as part of your analysis - it might be good to have a helping hand. In these cases, you’re going to need a researcher.
Once AI has been developed further to be able to handle theming, sentiment analysis and other key qual analysis techniques, it can help with the bulk of analysis from online live chats, online research communities where the volume of content is staggering.
But then the same issue will emerge with qual that did with quant. You’re still going to need someone to go through and make sense of the themes and sentiments emerging from the research.
If you have applied one of the basic principles of research (triangulation in particular), you’re going to have to look to a researcher to be able to blend the findings from all the different elements of the research to give you the insight.
Bringing the Story to Life
There might be some great data visualisation tools available that give you a quick glance at the data in charts and tables, however, this links back to the point I made earlier about giving people reams of data to sort through; this isn’t what clients want and the insight isn’t going to jump out at them.
As I’ve discussed in one of my previous blogs making research findings visually impactful is key in order to get the message across to the end user. Without it they won’t know what the key take-outs are.
How do you get this? The story of the research needs to be brought to life. You need someone who has been through, asking the right questions about the right part of the data, delving into the right bits of the research and then has spent the time to present it in a visually impactful way to get the true insight out and unfortunately. AI can’t provide this at the moment.
You Can't Replace Experience
Having an understanding of the business and the challenges they face is something that’s invaluable for adding insight to research findings. Being able to see the problems, apply the insight and tell the business what it really means for them is something that only an experienced researcher can achieve. Whilst AI might be intelligent and be able to save time for the researcher, I can’t see it being able to replace their knowledge and experience a research has gained over their career.
At every point in this article I’ve come back to needing a researcher to be able to bring out the insight from research. AI can help with elements of the analysis process but for the insight to be delivered, you need a researcher’s brain.
If AI can be incorporated to make our lives as researchers easier by doing some initial number crunching or theming and sentiment analysis of your qual, then great; but there is a time and a place for a researcher to take over, ask those questions which automation can’t and deliver the insight. How do you think artificial intelligence will impact the market industry, and how do you believe the role of researchers will change over the coming years?
Discover the differences between market research and data science in our free infographic. Or watch a short two-minute demo of our online market research platform to see our software in action.