Why workers resist data (and what Lenin might say)
Donald Farmer sees data adoption as a political challenge, not just a technical one
Today’s post is the second in a series that explores a persistent set of questions I’ve thought about while in the data industry: Is there a widespread resistance within business to the use of data? If so, what can we do about it? Following Mark Madsen's reply last Tuesday and ahead of Scott Davis's this Thursday, this post turns to Donald Farmer.
Over his career, Donald has helped shape two eras in data. At Microsoft, he helped SQL Server overtake Oracle as the world's dominant database, while leading what became the Power BI suite. Then at Qlik, he led the team that produced Qlik Sense. Since 2016, he's run TreeHive Strategy, advising companies on AI and analytics from what he playfully calls "a unique perspective" – literally, from his treehouse. He is a sought-after strategist and speaker, and he's made a career of turning technical vision into market-moving products.
He’s also the author of a book that’s much more interesting than the modest title suggests: Innovating: A short guide to making new things.
Here’s Donald.
Why do people resist using data?
Resistance: Reluctance or opposition to integrating data into decision-making processes, manifesting as skepticism, avoidance, or active pushback against data-driven methods.Why? I am guessing three reasons, in reverse order of importance …
Fear of Exposure: Data may reveal inefficiencies, mistakes, or biases, threatening job security or status.
"We’ve always done it this way” … decision-making based on intuition, hierarchy, or tradition is deeply ingrained in too many organizations.
Most important … Risk Aversion. Data-driven decisions can feel impersonal or counterintuitive. A reluctance to take a course of action, just because the data says so, rather than using their own industry knowledge, experience or analytic / synthetic thinking.
What is to be done? (As Lenin wrote.) Here are 6 bullets from some slides I use on this topic.
As Lenin wrote? In such a buttoned-down list, this stands out like white socks under a formal blue suit. As such, it should be taken as a statement, and examined.
So what did Lenin write? Among many other things, he asked why “workers” failed to develop political consciousness, roughly comparable to today’s workers failing to learn data skills. Perhaps it’s because today’s workers see data not a means of relief but a means of greater enslavement. The only improvements they can imagine are incremental, making what they do already just marginally more efficient. Data, they may suspect, is another trick by today’s bourgeois intelligentsia, the infamously well-compensated executives who wander through the lower floors now and then.
Donald’s reply to “what can be done” offers credible means to convince them that data can bring meaningful rewards. He describes tactics that would actually appeal to me — if undertaken with integrity.
Leadership Modeling: Executives should visibly use data in decisions and admit when intuition fails.
Celebrate Data Wins: Share success stories (e.g., "Using X metric increased retention by Y%”).
Reward Data-Driven Behaviors: Tie promotions, bonuses, or recognition to data usage (e.g., "Best Experimentation Award").
Long-Term Roadmaps: Link data initiatives to strategic goals (e.g., "Reduce customer churn via predictive analytics”).
Embed Data Teams: Deploy analysts within departments to co-create solutions (e.g., marketing analysts shaping campaigns).
Feedback Loops: Regularly ask teams, "What data would make your job easier?" and iterate.
These are all potentially credible tactics, at least based on my long experience with employers and my own reactions to their various tactics. These six ideas acknowledge implicitly that you can’t feed people data and apps and expect them to develop data consciousness. The effective leaders I’ve observed have envisioned, modeled, and rewarded, behavior that calmed my only moderately cynical view.
There’s only one likely stinker: the feedback loops. This sounds like one of those workplace surveys that now and then appear and disappear with no visible benefits. How can we be nicer to you? Do you want donuts at morning breaks? How about after-work feel-good discussions? Lunchtime jazzercise, perhaps?
The other five tactics aren’t promised, they’re just delivered or not. “Feedback loops” starts with a promise of action. Any delay at all, or any result deemed cheap in an already wounded workplace just feeds doubt. The bourgeois will either iterate credibly on “feedback loops” or risk seeing all six tactics — if they’re perceived as part of one package — burned alive in cynicism.
I think Donald is correct to see data adoption as political, as Lenin might have written if he were here. So the remedies must also be political, which Donald’s seem to be.
I agree with Scott -- but one point I'd add is that maybe part of the issue is that many folks have evolved based on a series of core values that have become constants and standard over their lifetime. To the point where there are a series of "sacred elephant" ideas that they simply can't or won't argue or reconsider. Let me offer a few...
* Orange juice is a great breakfast beverage. It's healthy and filled with vitamin C.
* Eggs are filled with cholesterol and are incredibly unhealthy
* Margerine has polyunsaturated fat -- it's healthier than butter
These are simple non-numeric messages that were programmed into many of us as children that have affected how we behave, what we believe, and how we act. My point is that when it comes to advanced analytics or data analysis -- many people did not grow up in a household of doctors, mathematicians, or scientists. Many folks were not "brainwashed" into the importance of evolving an idea to a hypothesis to a theory. Many want to believe they know the answer because of a lifetime of experience or some special skill or sense. Or -- they simply think that studying data is a waste of time and not necessary. They think the idea is proof enough.
Politics is the craft of change management. Most of these data-resistance manifestations can be viewed as inertia -- stickiness within a present mode, inertia against a change of habit or practice or worldview. Fortunately, there is a well-eatablished body of knowledge for Change Management. It is deeply rooted in Psychology and Sociology. It can make a huge difference, if only Data Management and Decision Management professionals can take their eyes off the digits and refocus on the people. Political? Yep. Like the rest of life in this space we share with others.