OPINION2 October 2023

What does the evolution of search mean for market research?

AI Innovations Opinion Technology Trends

Increasingly, the interface for everything is search, and this has significant ramifications for insights professionals, writes John Martin.

Search bar

Our digital lives continue to evolve at breakneck speed. While we used to scroll through pages of apps and tap on an icon on our phones, we now swipe down and type the name of the app in a search box. And while we used to scroll through pages of television shows on our smart TV, we may now choose instead to speak directly to the remote. And so it goes on. 

The point being that we expect a search interface to be available – no matter where we might be within the digital landscape – and, little by little, it’s becoming the primary interaction approach for many tasks. Meanwhile, although Google looks set to continue to be dominant for some time, it is clear that large chunks of search are breaking off into other places and spaces. In our most recent data, for instance, we see heavy search activity in short-form video apps such as TikTok and Instagram Reels, especially when it comes to specific categories such as recipes and how-tos.

The introduction of ChatGPT has shown us that some categories of search are likely to disappear altogether, being replaced by tasks completed by AI agents. By way of example, instead of searching for recipes with particular ingredients, it’s easier to tell ChatGPT what you have in your refrigerator and ask it what you can make.

But what does all this mean for insights professionals and, more importantly, brands? 

Here are four key considerations:

1 ) Consider new entry points at the top of the funnel
First up, we must consider new entry points for customer journeys. The historical importance of search centred around it being a remarkably strong indicator for intent. Until recently, however, search was confined to a small number of entry points – largely the likes of browsers such as Google and Bing.

Now, these entry points are proliferating – from TikTok to connected television (CTV) and smart speakers – which is making it more challenging to collect the most relevant data points, as well as increasingly important to ensure that insights professionals overlay an understanding and nuanced interpretation of the different contexts surrounding each of these.

2 ) Ensure access to behavioural data
In order to observe and understand fast-evolving behaviours researchers need real-world, observed data. Without this, and as platforms and channels continue to proliferate, insights professionals are left with ‘finger-in-the-air’ analysis. Meanwhile, we must look holistically at consumer behaviours and avoid siloed thinking and making assumptions.

3 ) Remember that cross channel is king 
With the advent of CTV, mobile and streaming around a decade ago, data and usage patterns became very fragmented, making it very difficult for the advertising industry. There are learnings that insights professionals can take from this experience and which are clearly applicable to the world of market research – not least the fact that an omnichannel understanding of the consumer journey is more important than ever. 

Individuals choose platforms that are most convenient for their situation and for that moment in time. They don’t follow predictable journeys and behaviours are far less linear than in the past: take the fact that a lot of people do not watch TV at ‘primetime’ anymore as an example.

4 ) Obsess about engagement – as opposed to vanity metrics
If we continue to think in terms of web pages alone, we are confined to old ways of thinking. Similarly, download numbers, for app developers or brands, can be little more than vanity metrics. For every service or app, engagement is what matters, while the most significant ramification of all these changes to the world of search is that the linear approach to collecting data  – which some researchers would call a survey – is no longer as relevant as it once was. 

Alongside this, new technologies, including the likes of machine learning, artificial intelligence and large language models, continue to impact all industries, all of the time. Central to all of these technologies is data, while exploring and understanding behaviours – what people are doing, watching, buying, and in what sequence – is what matters when trying to make sense of it all. Yet, despite this changing of the guard, only a small number are starting to excel by embracing new forms of observed, behavioural data. It’s time that changed, too.

John Martin is chief technology officer and co-founder at Measure Protocol