Will Analytics Displace Consumer Insights?
A recent article in Research Live - http://bit.ly/Hd3mKR - higlighted the rebranding of WPP's Insights Division from "Consumer Insights Division" to "Data Investment Management". CEO Sir Martin Sorrell explained the move as a way to bring the Insights and Media divisions closer toghether, thereby making it easier for Clients to manage and synthesis various data streams.
Eminently sensible move - but why rebrand, removing the word "Insights" completely, with the new entity sounding more like a financial services offering?
Has Insights as a concept passed its zenith? Are we in a world of "data" where the hot topics of data-mining, predictive analytics, synthesis tools will dominate?
I hope not, but if WPP is doing things, we need to take note. Here's my take.
1. "Insights" Still Needs to Earn The Right to that Name
How many Researchers have really made the leap to a value-added, action-oriented professional?
A recent Linked In discussion suggested that the switch from MR to Insights was little more than a change in name, with no significant change in MR practices, no gear shift in heightened impact. This was from someone who had spent many years in a senior Insights role at a major CPG company.
If Researchers are often still data-deliverers, re-branding Insights as "data investment" won't help much. Data Synthesis requires a high level of consultancy skills and business understanding.
2. Analytics is a Sub-Segment of Insights, not Vice-Versa
Primary Market Research is on an interesting, polarised trajectory, with two clear growth segments evident.
One is disruption-driven - more research conducted across the organisation and expanding to new industries that previously couldn't afford research, using low-cost DIY tools possibly in combination with in-house databases.
The other is the rise of more sophisticated techniques - biometrics, implicit questioning, techniques shaped by Neuro-Marketing and Behavioural Economics - aimed to capture more accurately emotional/intuitive as well as rational response, often on a longitudinal basis as with MROCs.
Alongside these, there is "Analytics" - utilising data-mining tools to interrogate existing datab to answer business issues.
Analytics is the more sexy issue of the moment, and no doubt there will be an increasing number of well documented case studies over time about how this improves business efficiency and decision making. Many companies are currently in proof of concept mode, figuring out the ROI equation.
However, the most powerful insights are invariably simple ones. They don't require a major consultancy to identify them through a sophisticated algorithm. They sit just below or actually on the surface, but can (and do) get overlooked for various reasons - myopia, cultural bias, political. Talking to employees will often give a good understanding of what both the challenges and likely solutions are to any given Company's problems.
Insights' task is to fight for these simple but powerful consumer or employee-driven "facts" to be acted on - these are ones that are likely to have the most impact. Analytics can support this, but certainly not replace it.
3. Algorithms have Way To Go
Ray Poyner recently pointed out some weaknesses in Big Data's (BD) ability to deliver - http://bit.ly/1810vLF - including the lack of people with BD skill sets, masses of meaningless noise as opposed to real nuggets (signals), no accepted models about how to treat BD.
Jannie Hofmeyer, on the other hand, warned recently at the MRMW 2013 Mobile Conference in London of the dangers of underestimating the power of machine intelligence when coupled with human expertise.
Both of them were touching on different aspects of the same theme: the ability of machines to replace humans. It's a core issue for Market Research - what will require our human skill sets in future, and what can machines simply do better?
My personal experience is that algorithms have way to go, as does machine intelligence. I recently took a year's subscription to a high profile UK magazine. The "welcome" letter had clearly been written by a machine, even though it was apparently from a human. It was addressed to me, but then referred to me in the letter as if I was a third person. Plus - the singular slipped to the pural. I quote:
Dear Mr. Appleton.....I can confirm that their subscriber number is..
Hallo? Welcome? To what planet? To continue my anecdotal evidence: my Facebook timeline has recently got peppered with ads from categories I have zero interest in. Amazon - surely a Big Data expert - often suggests I buy another CD from an artist I just bought one from. Duh. Cross-selling? It certainly has the potential to make me cross....
Overall, I see Market Research being sucked up into a fascination with numbers, predictability, machine learning, quantified cities and the self to name just a few. Digital technology is driving change, and fascinating us