Why research needs human intelligence, not just AI to succeed

Like so many industries, research has experienced extraordinary change in the form of new technologies and the emergence of AI and machine learning. But it is the human touch that will allow us all to successfully navigate this ever-changing landscape.

We now have more data at our fingertips than ever before and with AI and machine learning, people believe that you can solve anything at the click of a button. But can swathes of data produced by the latest tech guarantee greater understanding? Ben Korallus, Data Science Lead at Mintel doesn’t believe so:

“Today, we have more data and can use advanced statistical concepts to help us analyse this data, but it needs a human who knows what they’re doing to bring the insights to life.”

Toby Clark, Director of Research EMEA, adds: “It‘s about taking the data and explaining what it means, not just what the numbers are… The skill is about understanding people, looking at people and consumer behaviour and putting yourself in their shoes to uncover what motivates them.”

And this is something that only real people, not technology, can properly do. People understand people and we’re a long way off machines being able to reveal the nuances of human behaviour.

But in the world of research, the personal touch is not just an essential part of the analysis process. Our relationship with our clients now also demands a much more personal approach. While 20 years ago, researchers would write a report and send it off, crossing their fingers in the hope that their clients would read it. Today, it has become a collaborative journey; a partnership, with regular, ongoing discussions between client and analyst to really get to grips with the most relevant and useful data and insights to positively influence business decisions.

It is after all strong relationships based on trust that will allow everyone to fully benefit from the changes in the provision of data that are now a constant feature of our industry.

Mark Brechin, Director of Consumer Research and Data Analytics at Mintel explains further: “We’re at an interesting point for the research industry. There is scepticism about traditional research methods but also a reluctance to chuck out legacy data and embrace the changes. This is why the relationship between client and analyst needs to be built on trust and become a partnership.”

But building this kind of trust isn’t easy. One area that is sparking debate is just how important analysts’ understanding of research methodology is. Some believe that in a world where, thanks to AI, there are so many different data sources, understanding methodologies is simply a ‘nice to have’.

But Brechin strongly disagrees. He believes that a solid understanding of methodology is vitally important as it is exactly this type of knowledge that builds the confidence and credibility needed for a successful working relationship. Clark backs this point of view up:

“Clients will always query one methodology over another. We need to accept that no research methodology is the silver bullet – all have their merits but also their flaws. We need to understand these in order to identify the best option for the situation and know what we’re working with.”

Ultimately, in order to get the most out of the data you are working with, you need to properly understand it. As Korallus puts it: “Nothing can lie to you like a statistic can. If you don’t fully understand what’s going on behind it, you can gain the wrong understanding from that stat.” And it is a problem that will only be exacerbated by AI and machine learning as we continue to see ever more data thrown at people. “If you don’t fully understand them or you’re not applying them in the right way, they’re just going to lie to you better,” Korallus warns.

In the research world, we need people and we always will. We need people who can identify the very best methodologies and blend this with the latest data science to connect the dots and uncover the most valuable insights for our clients. As Brechin says: “Technology will only take you so far. You need people. You need analysts.”

To learn more, listen to Mintel’s ‘Little Conversation’ podcast, “The business of insights.”