The article shared below is titled "Will ChatGPT replace your consumer insights team". I doubt it. The point of the article is that the right balance has to be struck between AI and other research (market research) methods. Makes sense to me.
Data by itself is not an Insight. Observations by themselves are not an insight. Let's review some definitions or examples of "Insight" (sourced from the Corporate Executive Board's content presented at eBay a few years ago):
Example 1: Insight Is . . .
Reality - it is true.
Relevance - it is relevant to the brand or business.
Resonance - it resonates with the consumer.
Reaction - it will change behavior.
Example 2: "Insight is 3D. It's deep. It goes beyond what's said, done, and seen. It is directional. An organization can take action on it. It's 'Ding-Dong!' It strikes a chord with consumers - it provokes the shock of recognition."
Example 3: There are two parts "In" and "Sight". "In" is the psychological dimension . . what is inside the consumer's mind - casual motivation(s), attitude(s), belief(s), feeling(s). "Sight" is the behavioral dimension . . . what is observable - what the consumer is (or isn't) doing as a result of casual psychological factor(s).
You may recall the "Insight Pyramid". At its core is "Data". This includes what do we know? What don't we know? Information gleaned from many sources, including market research, sales reps, news articles, web searching, professional journals, etc. But data is not an Insight. Next up in the pyramid are Observations. This is the result of looking for patterns and connections in the data and identifying common themes. Here we start to understand why the data says what it does. But that's still not an insight.
At the top of the pyramid are "Insights". Insight is defined as: (1) Identification of some relationship or meaning . . . (2) with diverse sets of data . . (3) . . . that promises significant business impact.
Can AI do this? I doubt it.
I'm reminded of a presentation by Scott Cook, founder of Intuit. He was presenting on Customer-Driven Invention for New Offerings. Here's what I noted:
If you want to develop a new offering that's truly innovative, then (a) find an important unsolved customer problem that can be solved well, and (b) Solve it and build a competitive advantage.
His principles for finding the important problem:
Trust behavior . . focus on what customers already do.
Where's the problem's unsolved due to mindset (typically others have tried to solve a different problem or for a different customer)
The expected belief most in need of change may be your own and our own.
Seek and savor surprises and inconsistencies.
Get 1 to 1 with prospects: in-depth interviews & follow-me-home observation.
To summarize Scott Cook's advice . . . notice a clue in behavior, typically a surprise or inconsistency. Study behaviors 1-1 . . to discover important problem others have missed. See a path for the company/brand to solve it . . . that competitors are motivated to ignore. That was the birth of QuickBooks.
AI/ChatGPT is a tool. Insights teams and agencies will continue to flourish, in partnership with their clients.
Commentaires