Editor’s Note: The following article by Susan Schwartz McDonald, Ph.D. has been around for several months and has been well-commented within certain circles. If you are a regular reader of TheCustomer, you know by now that a part of our own manifesto is to search out relevant & valuable content and bring it to you – regardless of where it comes from. In this case, McDonald has penned an important and well-supported challenge to reform the insights industry. We think it deserves your attention.
To evolve our industry, let’s stop hawking vocabulary and commit to selling truly good ideas
There is much that ails our industry in 2019. Let me start with one big symptom. We have become bewitched by our own buzz-words, seduced by our pretensions. Every discipline uses words as window-dressing to sell ideas but ours seems to shill vocabulary itself. Consider how some of the mantras of the past decade have hijacked the conversation at various points. Terms like semiotics and heuristics, which flavored our speech without changing the storyline. Or neuroscience, which threatened to revolutionise the business but is now barely a whimper. The deafening drumbeat of behavioral economics, which has hammered precepts already native to the industry and others whose practical relevance remains proven.
Now comes behavioral science – touted literally by one industry conference as “the new buzzword you need to understand”. Behavioral science? The descriptive term that academic institutions use to reference their entire social science curriculum?
I don’t mean to sound petulant but here’s the truth: I’m fed up and I can’t take it anymore. We have simply got to stop being language brokers, who co-opt and repurpose words merely to get attention. And we’ve got to stop buying into words not backed by the full faith and credit of our industry. It’s hard enough to manage real change. We do not advance that process by turning phrases like a flywheel.
Who are we and what are we doing here?
If words are a symptom, what’s the disease? Living up to the overwhelming mandate of “innovation” — another word we hear entirely too often. There is, of course, plenty of real technology-driven change afoot, requiring new capabilities and new business models. And with extraordinary resources come extraordinary pressures. The call to do it better, faster, or cheaper (preferably all three) is making it hard for us to do it right.
The call to do it better, faster, or cheaper (preferably all three) is making it hard for us to do it right.
With this sense of urgency comes a strong presumption that if you’re not doing things in a different way than you did them yesterday, you’re obsolete, coughing in the dust of those who’ve raced past you. And if you can’t really change, you can simply rename.
We’re not just searching for words, we’re searching for ourselves. Vast, new data sources are allowing us to re-conceive research as the act of observing and overhearing rather than asking, which, in turn, fuels growing disillusionment with the basic premise of survey research: the idea that you can ask people what they think they think and expect to get a straight (or a useful) answer. As we refurbish our intellectual home with new kinds of data on a much larger scale, it’s not always clear where the old stuff fits in. Many data purveyors would like us to doubt that it matters why people do things if we can see so much of what they are doing instead. That’s not just change, it’s an identity crisis for the industry.
Homo Habilis. The best tools
No surprise, then, that we are mired in contradiction. For example, despite the avowed industry ambition to deliver strategic guidance based on synthesis and interpretation, our claim to 21st century relevance seems remarkably tool-centric. More real estate on company websites is being given over to specific “approaches” than ever before. Ironically, this deconstructive focus on tools can raise new barriers to data integration and coherent decision-making. We imperil the very thing we aim for – the best ideas – by distracting ourselves with buzzword-studded debates about the best tools.
And how we talk about things really does matter — in business discourse as in politics. Loose words can make it hard to get a firm grip on new ideas. Take machine learning. It’s an elastic term that can justifiably be used to describe certain types of regression analysis. As a result, it’s frequently invoked to make market research applications sound smarter and more believable than they really are. Even Data Science may be a bit of overreach for much of what we do. Reaching Mars with the Voyager takes real Science. Reaching authentic, bankable insights through a complex driver model probably takes more Art.
Honing our tools and our talk to spearhead real innovation
The Insights Industry surely needs to respond to this new world by innovating – but we’re unrealistic if we expect equally rapid and consistent advances in the statistical techniques used to explain or predict human behavior. Advances in brain science and behavior theory are proceeding at a slower rate than our data streams, and our models are struggling to keep pace with very rapid cultural adaptations that continue to thwart attempts at prediction. It’s always been hard to explain most of the variance in behavioral data. This frustrating truth is likely to remain true — no matter how much we beef up our tools with machine learning or we submit to artificial intelligence.
Perhaps, to be fair, that’s what our back-to-the-future nostalgia about old terms like behavioral science is really about. We are grasping at tools and terminology because that feels easier than doing the slower, heavier work of thinking hard about data to solve business problems.
Homo Sapiens. The Best Insights
We all know what motivated us to embrace that word “insights” — first as our rallying cry; then as the marquis for an organisational function; and finally, as the name for an entire industry. But with overuse, the word has been devalued. If we are honest with ourselves, we must acknowledge that today, any new thing we learn, including a driver model coefficient, can now qualify as an insight. Just as every small notch in our tool belt is billed as innovation.
As we reflect on ways to improve industry health and value, here are some things to consider. They are aspirational, to be sure, but so, after all, is true insight. The evolution of our industry hangs in the balance.
- We need greater transparency of ideas and clarity of terms. We won’t get far enough by hawking old terms or slapping cutting-edge words on ideas that have not necessarily been fully vetted. Real technical advances are worth crowing about — but we need to mean what we say.
- We need more research on research and more commitment to disseminating it. While there are clear challenges to advancing the cause of community learning in commercial environments, there are also meaningful opportunities for companies to partner and self-fund initiatives that lead to genuinely new knowledge. That includes collaborations between clients and agencies, and between agencies with complementary datasets and skillsets whose combined power can be realised only through creative collaborations. We also need to make better use of the platforms we have for exchanging ideas. We can advance best practices only by demonstrating value under the bright light of public debate.
- We need to ask unflinching questions about survey research in order to restore and safeguard what is uniquely valuable about it. To resolve the crisis of confidence in survey research, we must return to certain first principles so that we are better practitioners of the craft; and make a concerted, industry-wide effort to improve response rates and quality. This is an enormously challenging task but also an enormously important one. For certain critical applications, there is no substitute for hearing people tell us “what they think they think”. It would be a great loss to our industry if the next generation of insights professionals were to stop believing that.
- We need to develop a better understanding of the databases we aspire to work with. In too many client organisations, there is a gulf between those who manage databases and those who are responsible for wringing insights from them. This has led to poor communication and misunderstandings about what various databases contain and how they are structured. Things are no better (often worse) when we look beyond company walls to tap external data sources. If we want to do novel and ambitious things with Big Data, we need database literacy and full database disclosure. We ‘ve got to appreciate the concept of Big Noise along with Big Data. We can’t put faith in size alone.
- We need to rethink what we mean by data integration, and what we can actually do to achieve it. Data integration has become like the weather – something everyone talks about but can’t seem to do anything about. We mustn’t think about data streams as ingredients to toss into a stewpot and simmer with a pinch of AI. Once we recognise that various databases have structural misalignments, we must work to harmonise them – or conclude that they will taste better in different pots.
- We need better-educated, better-trained practitioners – drawn more broadly from the social sciences as well as business disciplines. B-schools appropriated responsibility for market research from the social sciences in the 80’s, but a two-year business program doesn’t allow for more than a nod to market research. We need to look again for greater thought leadership in the social sciences – not just by phrase-mongering, but by working to restore nourishing connections with the academic disciplines and organisations that seeded our development. It’s time to resurrect “consumer psychology” as a realm of serious study.
- We need to rethink the word, “insight” and re-energise it with intellectual curiosity. Insights used to suggest the sort of deep, glinting mineral vein of ideas about customers which, when unearthed, would make a powerful difference to our enterprises. With overuse, it has lost its luster. We can’t call the word back, of course, but we need to keep it vital by using it more sparingly. We must aspire to understand the “why” of human behavior, not just the “what” and the “what-next.” And we need to put ears to the ground everywhere, tapping cultural data from all sources as context for interpreting our research data. True insight is the product of synthesis and reflection – not just analytic automation
For all that to happen, we need to restore the word analysis to its former glory, right next to analytics, which has too often displaced it. Whenever I get airtime with young industry professionals, I extol the virtues of intellectual curiosity. Insight, like inspiration, favors minds that are truly prepared.