Using Teamwork to See Past Blind Spots – Insights From an NYPD Detective

By Andrew Grenville, Chief Research Officer | January 28, 2020


We all have blind spots, literally and metaphorically. In our eyes the optic nerves pass through the retina obscuring our sight. But we don’t notice because our brain automatically fills in the blank. When we look at research results, we also have blind spots—things we miss, connections we don’t make. And just like our literal blind spot, we can’t change that on our own. But it is a problem we can, and must, address.

Get by with a little help from my friends

We are not the only professionals that face this problem, and we can learn from other’s examples. Detectives, doctors, and the intelligence community all use some sort of peer review or teamwork to detect and confirm blind spots.

I recently interviewed an NYPD detective for a new book I am writing about how insights professionals can absorb fresh ideas from other sense makers. This detective, who must remain anonymous, told me the story of a car-jacking case he was working on, and how he used teamwork to help overcome his blind spots.

He now uses video footage a great deal because it provides concrete evidence—the kind that appeals to prosecutors and juries—and avoids the vagaries of eyewitness accounts. He said “They want to see the perp on video. So, a tremendous amount of what I do now is just pulling video from places. Just surveillance video, what have you. People’s cell phone videos, people’s pictures. Obviously, I work in Manhattan, so there’s video cameras everywhere.

“This case I’m working now, it’s a carjacking. I have these guys before the incident walking down 29th Street towards where they steal this car. So, now I just go back and check the video. ‘Okay, they’re walking eastbound down 29th. Now I have them crossing 2nd. Now I have them crossing 3rd.’ So, I work backwards, the opposite direction that they went.

“I tracked them as far as 7th Avenue, and I’m down here on the west side now, because they ditched this car on 26th Street. Then [the perpetrator] walks around three blocks circuitously and ducks into the shelter where he lives.

“So, now I’ll need to fight it out with this shelter and my department’s legal team to try to get someone to tell me who came in at 1:15 that night, because it’s obviously my guy who did this carjacking, because I followed him from 16 different cameras as he walked around Chelsea here and then into this shelter.”

But the detective doesn’t just watch the video alone. He asks others to watch too, because he is aware that he’ll be blind to some of what is going on in the footage and not know it. He explains “The best thing I’ve done is have somebody else watch it with me…We upload all these videos into our case files so people can watch them. People will review them and just email you and be like ‘Hey I reviewed your video, and it looks like the guy has hand tattoo. I don’t know if you saw that.’ Just strange stuff you wouldn’t think of. So, a lot of people watch these cases. The bigger the case, the more people get involved.” In the case of this carjacking he put extra effort into it because the victim was “really badly hurt.”

“It’s really helpful because they’re seeing it fresh,” he said. “If they don’t know much about your case, they don’t walk into it looking for what you’ve already seen. They see it on their own. They end up seeing more stuff, just other people, other involvement, or interpreting things in a way that you wouldn’t.”

Intelligence analysts work together too

Having one’s analysis commented on by others is also baked into the intelligence community’s process. “When your work begins to gel into something concrete, then [peer feedback] becomes a much more formal structure, and that has both a horizontal and a vertical aspect to it,” an anonymous former CIA analyst and current professor of intelligence analysis told me. “The horizontal part of it, before something goes to senior policymakers, is that it has to make its circuit around. Especially if you’re getting a little bit outside of your lane. So, if I would do something and it would have a heavy diplomatic angle to it, I would have to send it over to somebody at the State Department for them to do a sanity check on it. And certainly, if the issue was whether to deploy troops and military kinds of things, then it would have to go over to the Pentagon.”

The input from people with different sources of expertise is important because it helps us shake free of what can quickly become shared assumptions and group blind spots.

The input from people with different sources of expertise is important because it helps us shake free of what can quickly become shared assumptions and group blind spots. Intelligence Analysis for Tomorrow: Advances from the Behavioral and Social Sciences is a report by the National Research Council that recommends building processes in which feedback is provided by people from across an organisation. “When people with heterogeneous backgrounds work together, their perspectives emerge in different ways, allowing more knowledge and solutions to emerge.

Diversity can be sought in subject-matter expertise, functional background, personal experience and mission perspective. Such sharing allows analyses to be richer and deeper, with better understood strengths and weaknesses, whereas individuals working in isolation are more limited by their assumptions and myopic about the limits of their knowledge.” These examples of how detectives and the intelligence community deal with blind spots should encourage us to rethink our own analytic processes. The world of insights needs to do more to solicit ideas and feedback from multiple stakeholders. Too often we have isolated and myopic analytic approaches that seem designed to encourage tunnel vision. We must make better use of teamwork and peer review to ensure our insights reveal the full picture.

This article was originally published on ESOMAR’s Research World

Previous
Previous

Utilities and The State of the Consumer

Next
Next

What is the Price of Privacy in Healthcare AI?