Skip to content

A Common Trap That Undermines Analytics Credibility

Over the years, I’ve seen analytics professionals of all stripes blow their credibility and lessen their impact by falling into a common trap. I have to admit that I fell victim to the same trap early in my career. While our intentions are pure, our analytical minds and approaches can get the best of us and we explain too much. We’ll be better off if we learn to provide less detail and stop talking sooner than we are naturally inclined to.

Don’t Say Too Much!

It seems like helping a business sponsor understand all of the details of how we created an analytical process would increase our credibility. After all, who wouldn’t be impressed with all of the issues we handled, all of the modeling approaches we tried, and all of the validation we performed? Unfortunately, it doesn’t actually work that way. It is easy to understand why if we consider a visit to a local auto mechanic to deal with a broken-down engine. Let’s see what two different mechanics might tell us and then consider which we’d actually have more confidence in.

Mechanic #1:

“It looks like you have a transmission issue. I’ll need to run multiple tests to validate that including a fluid pressure test, an electronics test, and a heat assessment. Let me explain those tests to you. (… after 5 minutes of detail on the tests …) Depending on what the tests find, I may or may not be able to fix it quickly. One less common scenario I’ve seen with these symptoms requires an entire new transmission, which would cost several thousand dollars and take about two weeks. While that scenario is possible, it isn’t probable.”

“We’re also having issues with our main supplier right now. I don’t expect we’ll have an issue with transmission parts, but I just wanted to alert you that we won’t know for sure until we place the order. I can’t give you a firm time and cost estimate until tomorrow or the next day, depending on when my sick assistant mechanic gets back to work. Should I check the car in?”

Mechanic #2:

“My gut is that this is a transmission issue. Most likely with these symptoms I can fix it in 3 days for about $1,500. While I can’t commit to that until we take a look, I expect that’s what we’ll find, and I’ll be able to give a firm commitment within 48 hours. Should I check the car in?”

The Verdict

While mechanic #1 provided a lot more detail on the tests, risks, and range of outcomes, you’d likely leave that discussion feeling uneasy and overwhelmed. He told you much more than you needed or wanted to know. You wouldn’t want to leave your car with him. Mechanic #2 was concise and confident. He acknowledged that there were risks, but he stated what he thought the issue was and was very clear on the path to completion. You’d leave that discussion feeling much more comfortable than the first even though you were provided much less detail. Be sure you aren’t giving so much detail in your explanations of an analytics process that you become like mechanic #1. Keep it simple and confident like mechanic #2.

You Can Always Add Detail, But You Can’t Take It Away!

The moral here is to only go into details if asked. Most people don’t care about the details and only want to know that you can handle their problem. They have come to you, much like you’d go to a mechanic, because they are assuming that you have the knowledge they lack about how to address their analytics problem. Diving into immense detail doesn’t actually add to your credibility. Rather, it will tear it down as you plant seeds of doubt and add the possibility of mayhem into the mix.

Think of giving information like filling a glass. You can always add a bit more if it isn’t full enough, but once the glass overflows you have a big mess on your hands. Once the mess is made, you can’t take it back. So, always add information slowly until it is clear you’ve hit the level that is required, and no more.

A friend gave me a great analogy for this process. Only provide a tree trunk to start. If someone wants to understand the branches, then go to that level. If they then ask to understand the leaves, then go to that level. But, don’t go any further than you’re driven to. If you always drill down to the gory details, you’ll fail with all but those wanting the gory details. If you always start high level and only get more detailed as asked, then you can succeed with everyone!

Offer The Right Perspective

It is also important to make sure you’re properly interpreting the level of detail being requested from you. It is necessary to interpret the request from the view of the person asking.

Imagine going into a car dealership and asking how a backup camera works. If the salesperson starts talking about the super-high definition OLED screen, the onboard sensors that are scanning the area in real time, and how the trajectory lines are computed and rendered, you’ll probably wish you hadn’t asked. Very few people care about those technical details like a car salesperson does. Most people would just want to know that the camera will show what’s behind you, show where you’re heading with some lines on the screen, and it will beep at you if you’re getting too close. While the question may literally be, “How does this work”, what is usually really being asked is, “How do I use this?”

Similarly, if a business executive asks, “How do these analytics work?” the desire likely isn’t to understand the math, the data issues, or how it will be deployed. Rather, the desire is likely just to understand how to use the results. For example, a simple answer might be, “The analytics will provide a probability of success for each customer and you can then make a decision to act based on that probability.”

Check Yourself

It can be very hard for analytics professionals to hold back on the details that we get excited about. However, we have to remember that those we’re building our processes for usually couldn’t care less about those details. Our credibility isn’t helped by overwhelming business partners with technical details.

A client said it best. He said, “Share the What, So What, and Now What. Only go into the How and Why if asked to.” We actually serve ourselves and our sponsors best when we say the least we can to instill confidence that we have the situation handled and to provide guidance on what to do with the results. Be sure to check yourself and avoid falling into the trap of saying way too much!

Bill Franks, Chief Analytics Officer, helps drive IIA's strategy and thought leadership, as well as heading up IIA's advisory services. IIA's advisory services help clients navigate common challenges that analytics organizations face throughout each annual cycle. Bill is also the author of Taming The Big Data Tidal Wave and The Analytics Revolution. His work has spanned clients in a variety of industries for companies ranging in size from Fortune 100 companies to small non-profit organizations. You can learn more at

You can view more posts by Bill here.