With a glut of sophisticated products on the market, can AI-led research tools transform the industry?
The bold promise of AI market research is that insight gleaned from monitoring virtual focus groups, analysing customer data, scraping emotional content from social and the like will revolutionise the research industry. Lightning quick algorithms will save time and scythe through projects leaving researchers in a crumbled heap of flesh, bones and rumpled chinos.
In reality there is no silver bullet. AI tools can certainly help us speed up the research process. However, the key determinant of success, is less about removing the lumbering human from the process and more about how agencies are capitalising on the extra hours they can dedicate to projects.
Time spent thinking creatively can’t be crunched by technology. Faster doesn’t mean better quality, and time saved from using AI techniques should be reinvested into creative thinking to bring more value for the client.
There is a real danger with AI that the industry will expect insights to be found instantly, thereby enabling equally rushed decision-making. Instead of thinking of AI as a means to deliver faster results, a smart consultancy is the one which is confident enough to use the time saved to deliver more considered, strategic insight. This will lead to better, rather than faster, decisions.
As a customer consultancy we rely not just on primary research, but on immersing ourselves in our client’s problem. Capturing crucial research context is essential for clients – especially for big heritage brands looking down the barrel at fast-changing customer landscapes. Heritage brands need to identify the direction of travel and sniff out the next big thing if they want to keep pace with agile start ups.
Better quality analysis
AI methodologies can give us more information about our customers than ever before. But while we may be moving to a state where a couple of keystrokes reveals everything about everyone, it’s worth noting the same information is also up for grabs by every other agency. The point of difference for a customer consultancy comes from decisions driven by nuance, informed analysis and entrepreneurial imagination. It’ll take humans (the best ones too) to deliver on that.
Three AI methodologies that have caught our eye
Groupsolver combines elements of the focus group with online research to offer deeper responses to questions, fast. It works much like a group instant messaging chat, though one in which ‘moderators’ can pose questions. Anything up to 200 respondents can then give answers that the whole group can see. On the face of it, it’s a focus group with a robust sample. While it’s a compromise on the intimacy of a face-to-face focus group, GroupSolver promises a level of depth that moves beyond the quality of responses associated with traditional survey open ends.
Their software can analyse, and quantify, and spit out seamless outputs outlining the key themes coming out of the conversations, enabling seamless trackback to the ‘making decisions part’. It’s not hard to see from this how insights can be reached with plenty of time to spare. If you’re thinking “How much value can be garnered from short verbatim responses?” you’d have a point. This is clearly no replacement for the depth of understanding that a face-to-face focus group will give you. And it’s this depth that often leads to the most powerful insights. However, AI-based research methods need not replace traditional methods. They come into their own when they’re used as a tool to give a sense of the right direction within a limited time frame.
Discover.ai’s tool has the ability to analyse a multitude of online sources at rapid pace. The tool scans these sources, simplifying millions of words from them, and making the connections that ultimately produce an output of compelling opportunity springboards. These engaging springboards are a neatly presented and charted-up list of key customer trends and needs. Engaging quotes help bring them to life. These trends can be used as starting points for product development, or other innovation workstreams and are potentially very powerful.
Before we kill desk research entirely, it is still humans who select the sources that are inputted into the software – the software does not magically choose the ’best’ sources. And, ultimately, the AI created ‘springboards’ aren’t standalone insights. They are just the start point, inspiration for research teams on whom the responsibility lies to create the real edge.
Those who remember the wild years of the early 2010’s will know that dense data packs or dashboards can be overwhelming. Text commentary is often the most effective way of engaging stakeholders who may not be as data savvy as the insight team. However, creating these is time consuming, which is where natural language generation (NLG) can help. Wordsmith: automated insights is an NLG tool that automatically transforms structured data into written narrative, rapidly speeding up the reporting process. Quick though it is, it is not a completely self-sufficient process. A researcher has to set the narrative design or the ‘rules’ to ensure that the powerful NLG software pumps out a suitable narrative structure.
The human touch
There is a temptation to see AI as part of a clinical, automated, objective process that arrives at the answer at the click of a button. Scratch beneath the surface, and it quickly becomes apparent that human fingerprints are all over it. We still have a long way to go before we reach the stage of the game where AI reaches compelling conclusions without human guidance. Much like a skilled qualitative moderator, or a quant questionnaire expert, AI-led research still needs intelligent and curious consultants.
AI’s ability to save time on a project is immense. However exactly how consultants choose to spend that time, more so than the methodologies themselves, is what has the potential to transform customer understanding into strategic dominance for clients.