Note: This article first appeared in the September - November 2020 edition of the Research Society’s publication Research News.
Back in 2010, could we have predicted the state of the world and our industry in 2020?
Some trends like the accelerating migration to online research and panels from face-to-face and telephone interviewing were predictable. The exponential rise of Big Data and machine learning was always on the cards. The waft of privacy concerns was in the air based on the rise of social platforms, but pre ‘Cambridge Analytica’, I don’t think people really understood the implications of customers becoming the product.
Instagram started in 2010, as did Quora, Flipboard and something called Chatroulette. Read into that what you will. Predicting the future is a mugs game indeed. Let’s start with the ‘known knowns’ as of right now.
Time is precious, surveys are cheap and everyone wants to know something.
People don't want to do traditional surveys. It's getting to the point where even conducting a government census is hard. Pollsters can't even pick elections anymore. People are happy to share an opinion if they care about the topic but won't answer the phone from a stranger. They'd rather just jump on reddit and get upvoted or announce their authority via an opinionated Facebook post…
Privacy is already a huge focus globally, and it can only become more important in a world where our faces can be recognised in a crowd, our speech tagged and behaviour predictively modelled. Which brings me to the next point…
We have more and more computing power coming online – we can model things to astonishingly (creepy) accuracy.
We can create fake humans on screen and actors are appearing in films after they die. We are getting closer to passing the Turing Test (faking a human) with believable facsimiles of machine human interfaces.
Yet, with all this technology, productivity growth has actual slowed in the last decade. It is getting harder to squeeze more productivity out of the economy. Maybe that's because we spend all our time on social media being opinionated. Organisations are trying to move at a faster and faster pace and ‘fail cheaply’ but marginal gains are getting harder to find.
Back in 2010, touch screens were ‘wow’, but devices are now clearly moving towards voice interfaces. Menus and tabs are so yesterday. Voice is in your car, your home - suddenly talking to a block of plastic is normal.
So, where to for our not-so-little $80b+ world of market research? Let’s focus on some reasonable extrapolations:
Prediction 1: Conversation > tick boxes
People like expressing themselves in words – written or verbal. That’s why Conversational AI is the next generation of insights. Why ask someone to rate something, when you can tell by their tone of voice or choice of words? This is truly exciting and better for both businesses and consumers. You can tell the bot to #*&! off and it won’t even get mad – just empathise. This technology is improving exponentially and for the first time we have the analytics to ACTUALLY make sense of it. The amount of quality text information you get through Conversational AI is 275% greater compared to a structured survey.
Prediction 2: Don’t do another survey, just improve your model
If you pile enough data into a neural net, you can model the world. If you have a world model, you don’t need to do research, you can use your model to predict a result and it will give you the answer in next to no-time. The company with the best world model wins. To get a sense of how quickly this is moving forward, read this article on a language model with 175 billion parameters.
Prediction 3: Insights on the desktop
The days where a research process must be managed by a team of people are coming to an end. Great systems deliver insights in real-time on demand; they answer questions, interact with people at work and support decisions. Think ‘decision support systems’ powered by sophisticated models and delivered by AI voice interfaces at work.
Prediction 4: Guided insights, not dashboards
Dashboards will remain important, but you won’t have to search what you’re looking for – it will be delivered to you before you know you need it. “Bob, you need to know that satisfaction is trending down because people are running out of toilet paper and can’t wipe their butts’. ‘John, it seems there are some people in New York are talking about activated almonds. This is also trending on Reddit’. Join the dots, spot the trends, react faster, get ahead.
One more prediction for the road - We are realising collectively that our planet’s resources are finite, and that continuous GDP growth is unrealistic. The majority of market research is about growth and finding ways to sustain it. My hope is that in 2030, we are spending more time understanding how to live sustainably and develop ideas for doing more with less and being happier without consumption. Now, there is something that makes me excited.
