How many times has this happened to you: You get to the end of a gruelling analytics project. You’ve tortured the data until it confessed. You’ve made a silk purse out of a sow’s ear. You’ve found the fountain of youth. Against all odds, you solved it and know without a doubt that you have the right answer. You are thrilled. All that effort was worthwhile.
And then you present your results to the decision-maker and he disagrees. He questions the applicability of your data. He doubts the accuracy of your analysis. He remains unconvinced even in the face of overwhelming evidence. He rejects your recommendation and chooses a different alternative.
You console yourself with platitudes:
“You can lead a horse to water, but you can’t make him drink.”
The result, of course, is project failure.
This situation plays out in organizations every day. Thousands of analysts do brilliant work, only to see it questioned, ignored, and ultimately, rejected.
So how do we fix this? Are analysts destined to flock to the same set of analytics-friendly managers? Is your company doomed to mediocrity because the decision-maker hates numbers?
Before we answer that, let’s turn the tables and look at this from the perspective of a decision-maker:
You have grown up in an industry and been promoted over the years because you have a track record of making good decisions, developing good people, and growing the company.
You have devised a number of time-tested shortcuts to deal with uncertainty. You’ve surrounded yourself with people you trust and you weigh their advice before making major decisions. You put your reputation on the line with every one of these decisions.
Now, perhaps you have an analytics guy. He’s new to the organization, and when it comes to decision-making, he approaches it very differently from your other advisors:From time to time he suggests a solution to the wrong problem. He hasn’t grown up in the business so he can’t be expected to see the big picture, but still, he seems to have a hammer and so everything is a nail.When you ask him for an opinion, he says “I’ll get back to you” then disappears for a month or two to work his black magic.When you ask him how he arrived at his conclusions, he begins to speak a different language. You sometimes wonder if he even understands the words he is using.He keeps talking about his models as if they’re some kind of authority: “The model says so”. But he has a hard time explaining why it says so.When you question him about the data quality, he waves it away and mumbles something about sample size.Whenever he submits a report, it reminds you of that 4th year econ course you almost failed. He points to numbers and charts that should be understandable but aren’t.
In short, this analytics guy is a little bit weird. He surfaces from time to time with ideas and opinions, but he can’t justify them in a way you understand. To make matters worse, he seems offended when you don’t take his advice wholesale. You suspect he won’t last long in the organization.
This story plays out all the time. On the one hand, we have the analyst providing objective, testable truth. On the other, we have the manager drawing on years of experience. The two perspectives seem irreconcilable.
But they aren’t. There’s actually a simple solution, though it’s not one you want to hear. You see, as the analyst, you are the problem. Defining your role too narrowly is preventing you from contributing.
You have to own the adoption. You have to accept that the job isn’t done when you get the right answer. It’s only done when the decision-maker is convinced.
You heard me correctly. No more blaming your pig-headed boss for ignoring the gem of an employee you are. If he’s not accepting your advice, then you’re not doing your job.
This sounds patently unfair. And it is unfair. But if you adopt this philosophical shift, it will inevitably lead you to change your approach to your job. It won’t just affect the way you present your work. Instead, you will end up thinking of adoption from the very beginning:
You’ll spend more time up front understanding the problem. You’ll put yourself in the decision-maker’s shoes and ask more questions. You’ll get the problem stone cold right.
You’ll provide more frequent updates. Instead of the big reveal at the end, you’ll provide incremental results. You’ll treat the decision-maker as a resource to improve your approach instead of a roadblock to remove.
You’ll choose simpler models. You’ll be thinking about how easy it is to explain what you’re doing. You will come to value clarity more than an additional digit of precision.
You’ll practice your storytelling. You’ll teach yourself how to translate geek to business and not just business to geek. You’ll use rules of thumb and narratives to explain both how and why your method works.
You’ll visualize way more. The eyes are a shortcut to the gut. You’ll bypass the decision-makers inherent doubts and connect right to his instincts. Finally, your charts and tables will be clear and engaging.
The quality of your analysis will benefit too. You’ll be continually poking holes in it and you’ll be much sharper because you’ll know that it’s going to be used (and scrutinized).
In short, you’ll no longer be a resource person providing the “analytics” on a particular decision. You’ll be embedded in the entire problem-solving process. This change will benefit both you and the organization.
The biggest hurdle you’ll have to leap is your own sense of injustice. Leap it. Own the adoption.