What is dark data, and how do we combat it? In this episode of The Difference, the podcast of sister company Brain+Trust Partners, listen to Trust Insights CEO Katie Robbert as she discusses dark data, organizational theory, and how effectively managing the people creating the data is the fastest path to lighting up dark data.

Don’t see a player? Listen on Soundcloud.

Machine-Generated Transcript

What follows is an AI-generated transcript. The transcript may contain errors and is not a substitute for listening to the episode above.

There’s a wise and admittedly often use quote by one Sherlock Holmes that speaks to one of the biggest problems facing executives today. It goes like this. “You see, but you do not observe that distinction is clear.”

So often on this show we’ve talked about the difference between the incremental and the fundamental, the superficial layers of our work, and the foundation itself. Big brand executives who are able to successfully transform their organizations for the digital age. They seem to be fiercely obsessed with the ladder with a fundamental the human element of what goes on in the world. They understand what many don’t seem to even notice because while most leaders see, they don’t observe. You see there’s a distinction. You don’t need data about your customers you need insights.

Recently, a new company was born called Trust Insights.

It’s a connected venture with the makers of this show, Brain+Trust Partners, but just to clarify, they are two separate entities capable of bringing on separate clients or collaborating on the same client. And today we’re going to talk to the co-founder and CEO of brain trust insights Katie Robbert and specifically, we’re going to talk about this crucial distinction between seeing and observing. The problem; dark data. The solution, well, it’s all about shining a light on this buzzy trendy ubiquitous concept in our world. So do you see the world around you or do you truly observe it? The distinction should be clear, but just in case, let’s explore the difference.

Welcome to the difference the podcast from brain trust partners. I’m your host today, Jay Acunzo and here’s something rather, well, dark. About 80% of the data that brands gather is unstructured. In other words, not readily readable, or understandable or useful. Unlike the data captured in say website forums or Google Analyt,ics many other sources leave you with unstructured data these sources or things like customer notes left by customer service reps inside of companies call center or the social media comments that are brand receives after they post something or because of it. This could also include customer reviews left on retailers product pages or video data or image data and of course now voice, data. There’s a lot of different ways that you get inputs from the world that aren’t readily readable and useful as data. And worse, you can’t find insights from these unstructured sources. According to market intelligence firm, IDC 90% of this unstructured data is never actually analyzed this is like being a detective gathering bunch of clues and never even using it. You just stash it away in an office somewhere.

Luckily today. Katie Robbert is on the case:

Quite a few years back I was a product manager at a health IT company and one of my charges was to present information on a monthly basis to my steering committee. It was comprised of about eight different stakeholders from all the different disciplines. So you can imagine that they all came at it from a different perspective from sales and marketing to development to UX to the C-level suite and you know everybody had a different agenda. My primary role was to present them with information to get them all to make a singular decision which, you know, truth be told it didn’t always go well because, you know, at any minute. Someone could Blitz and say you know what I’ve decided I’m going to change my agenda and I really care about this thing over here. Right people, people weight. Exactly. So, you know, it was always a challenge, but it was a challenge that I really enjoyed sort of in a weird too.

It was like okay, my mark of success is if I can get everybody in the room on the same page with the information I present to them. I was only competing against myself but my personal challenge was to present to us to collect and present all of the information in such a way that I could get everybody on the same page. And for me, those were sort of like my achievements that I would be seeking out. And you mentioned before you use like a signed document or something like it sounded like you were basically like insanely to just tell me a little bit more about that.

I was and still am insanely rigorous about details because one of the things, and you’ve probably run into this as well, is you know getting someone’s verbal agreement, especially on a big decision isn’t good enough. So I would i would basically document out all of the decisions that I was hoping that they would you know this at the steering committee would make and then as the decisions were made, I would literally stop the meeting and have them or at the end of the meeting.

Like initial all of the things and sign it and data so that when like two weeks later, my health care you know stakeholder would come back to me and say, actually, I think we’re going to make it blue and not green I could put the paper in front of his face and be like: Well, actually, on Tuesday at 3:27pm. It was raining outside you were wearing a purple shirt. You said we were going to be green not blue. And it became one of those things like you can’t argue with the data. And I mean one of the other things that people really don’t know is that I also have a paralegal degree and so I’m really good at documenting and making sure all of the details lineup and make in like so if you need to go backwards in time and try to figure out what happened, I’m the person who always has that information.

What was the response from people while you were doing this? I feel like as an executive I get up to a certain threshold and my career because I’m the one who knows or things I know something. Correct. And here’s somebody else a product manager, perhaps, who’s saying sign this document. By the way, I’m holding you accountable and oh my gosh how many people in business hide from you know saying what they mean and meeting what they say out of fear that there’ll be held accountable and you’re doing exactly that. So like, How was that received oh they hated it.

They hated it. It was not a popular way to do things. But ultimately, at the end of the day. My job was to get them to make decisions and stick to those decisions and if that meant like holding them accountable with this really retentive lens of detail than that was what I needed to do because it was in the best interest of the product A customers and the business because a lot of the decisions that they would make we’re tied to money so you know if they wanted to make something blue instead of green I would need to be able to present to them. Okay. But that’s a $20,000 decision. So everybody here on this day is making the same singular decision if you change your mind you’re going to waste $20,000. Is that what you want to do?

And so I had to take my own fears and ego out of it and say, You know what, this is my charge. My goal is to get you to make a decision and stick to it with the data that I am presenting to you. So yeah, I got a lot of pushback. I got a lot of flack from it. You know, I probably got spoken to a lot about holding them accountable. But at the end of the day, it was what made me so successful and what made the product so successful with the customers because we were giving them what they asked for versus what the leadership team thought customers wanted. So tell me about dark data. I feel like this is a phrase is become a little bit of a buzz word for some but also something that’s been elusive to understand for many so what how do you define dark data, you know, it is funny because it is I think it is becoming a bit of a buzz word but at a very basic level dark data is data that you own.

That you’re not doing anything with. So a really great example is if you have your Google Analytics and your car and your marketing automation and your customer support systems all collecting data about your customers you don’t necessarily know what all of that data is saying all in one place. So that’s what you consider your dark data because it’s just data that’s baffling to you. It’s data that you’re not using it stated that you haven’t looked at, you know, you could use a very simple example as well.

Okay, so I wear a Fitbit every day. But I don’t necessarily look at it. So why am I collecting the data, you know, I’m not gonna it’s not data that I’m going to necessarily do anything with it’s just kind of there and just kind of collecting all the time. For some reason, whenever you talk about dark data when we’ve spoken in the past, I go right to big data in my head is like this giant trend. It’s like there was a moment in time where everybody seemed to stand on the roof of their buildings and run around their offices screaming before about big data without ever stopping to wonder like what is this doing for our business. So, you know, did that obsession have an effect on this problem of dark data as it related somehow.

Oh, I think so. Absolutely. I remember again back at that same health IT company when a lot of my senior managers just started saying the word big data, and it really didn’t mean anything. They were just saying it because they were all of these articles from, you know, Forbes and a bank and Harvard Business Review coming out about big data. So they were like, Oh, well, we need to be part of big data. And the first question is, well, what does that mean and what does that mean to us as an organization. And so I think, you know, dark data and big data definitely sort of have the same that sort of elusiveness of like once I know what it is it’s going to solve all my problems. But honestly, it’s not that mysterious Big Data is you know again at a very basic level, just a lot of data collection. It’s very large datasets. And then dark data is data that you’re not using to make actionable to bring in more customers or make more money or not lose money.

Here’s a quick example of a company that shed some light on it starts data. In other words, they were able to analyze Android insights from a treasure trove of information that was previously untapped this example is from Stitch Fix the online subscription shopping service. According to an article published by Deloitte Stitch Fix uses images from social media and other sources to follow new and emerging fashion trends among customers and then personalize their experience they start their process with customers with a lengthy questionnaire about their clothing preferences as they sign up if you were to sign up yourself, you’d see a bunch of new photos about new items and new fashion trends and then you tell the company, how you felt about those aesthetics. And then lastly, you would have the option to link your Pinterest account to Stitch Fix And oh, by the way, the company knows that their customers actually use Pinterest quite a bit, so.

All of that data is unstructured and quote unquote dark unless it’s in a survey that you fill out. So all the imagery is somewhat directional but not overly useful, especially at scale until they have that data to a team of 66 zero 60 data scientists working for the company to then analyze the data and get a better sense of a customer’s style. As a result, the company’s stylists as well as some proprietary AI algorithms that they’ve built create a more relevant box of clothing items shipped to you every month and it’s absolutely crucial that they get this dark data and then unlock it, they structure and find insights in it because it can be really hard to capture structured data about something as nebulous and also difficult to explain as fashion or style trends. This is a quote that Julie Bornstein the company’s CEO told digit day and Julie, by the way, is the former CMO of. So for us, so she knows a thing or two about style and fashion. Here’s what you said style and fashion is so nuanced and words can mean different things to different people going digital replaces that moment when you walk into a store talking about the customer and try to describe to somebody. What’s your closet looks like and can’t form it into coherent sentences. So in lieu of those coherent sentences in lieu of great beautiful language from customers to describe exactly the type of fashion that they’re looking for, they have to go into the world of dark unstructured data they have to capture that image data from Pinterest and from their own website now Stitch Fix essentially cells style but not just any style. It’s not prescriptive style. It’s your style. They have to personalize the experience and to do that. The your part of your style needs to be easier to understand and it needs to be analyzed in a repeatable scalable fashion. The end result for the company is a far more personalized experience with the brand which as we explored in Episode One of this show and simply not the norm. When it comes to most personalization efforts by companies. This is just another example of the power of shining a light on dark data. So take solace in the fact that yes, my dear friend, he can be done.

What makes you mad about the companies that get this so wrong because like you said, it’s not doesn’t have to feel like rocket surgery, especially the approaching it with the right mentality doesn’t so how you know you’re fired up about this topic in general and the opportunity and that’s the positive spin. What about the negative stuff you see happening, what really you off and grinds your gears?

What grinds my gears? You know, I think what it is it’s if the efficiencies that are lost. It’s the money that’s spent you know it doesn’t have to be something that’s really overly complex and difficult and that’s one of the things that Trust Insights really wants to bring to the table is we can we can take your complex spider web problems and make them really simple solutions and actionable things.

And so I think that is sort of what irritates me when I go into these conversations with other organizations is they’re overcomplicating it. It doesn’t need to be nearly as complicated as people are making it. And that’s sort of the legacy of well we’ve always done it this way. Okay, I was gonna ask why what’s overcomplicated like why is that happening you know companies tend to invest in a lot of different systems that you a lot of different things. So like I said, sort of that tech stack of you have your CRM and your marketing automation and your customer support and your analytics. And so, you know, a lot of times, all of those systems are necessary but what we see is that people don’t know know what to do across all of those systems to make them all either talk to each other or when you pull out the data to make them tell the same story or to all be part of the same story and I think that that sort of what I find I don’t know if infuriating. So right we’re but it’s just sort of frustrating to see all of all of the money being spent with software and all of the money being spent you know to collect the data that you’re not doing anything with and it could probably be a lot more simple even sort of you look at the larger companies that are so silos. The left hand doesn’t know what the right hand is doing. There’s teams that should be working together that don’t because that’s the way they’ve always been structured and the bigger the company, the more complex that gets one of my one of my friends and marketing likes to say if you see a ton of numbers. That’s a data dump. If you see plain English. That’s an insight. So how do we go from just a data dump to an actual insight, so you know how you get from a data dump to, you know, plain English insights is you need to be curious about it. You can’t just say, Okay, well this spreadsheet says five. So the answer is five. You know, sometimes the answer is five. And sometimes it’s not. But if you’re not willing to at least explore more and confirm it. You know, then that’s just an assumption and, you know, you tend to get into that rut, sort of, as you said of like people just sort of chugging along doing the same routine over and over again and what really gets you to those insights. It’s curiosity. Well, why is it five How did it get to be five what factors, didn’t we look at that brought it to be the number five. What is the importance of curiosity, really, like, why is why does that matter so much I think too often when someone takes a look at a data set, or someone says, Well, our customer said we’re doing a great job. So that’s it. I’m not going to dig into it any further, I think curiosity is sort of that ability to accept the good in the bad so you know if you dig deeper into your data, you might find out that you’re actually not doing that. Great. And that could be a scary thing and it might mean making significant changes. And I think that there’s a lot of people and a lot of companies that would rather just sort of like have that surface level assumption of well, nobody’s down our door and telling us that were terrible people. So we must be doing fine.

Let’s hear now another example of a company unlocking better insights from previously dark data. The last example Stitch Fix was an important one, but ultimately pretty fun. It had to do with fashion. So let’s talk now about something more serious and life changing Child Welfare according to North Woods a software company focused on human services agencies, the digital age has completely overwhelmed, social workers, not too dissimilar for many of us at argue, but here’s one stat that is indeed overwhelming. The typical child welfare case can include up to 5000 pages of information 5000 pages about that one child and the surrounding case and much of it. According to North Woods almost 80% of it can be considered dark data, it’s hidden or impossible to retrieve.

Therefore, can’t be used by the agency when making safety decisions for children. This problem can extend to agency wide trends, like how the opioid crisis is impacting foster care or case specific information that could prove vital like quickly identifying the people connected to a child like relatives or other positive influences that might be able to take that child in if they’re in danger. So use cases for shining a light into that dark data range from social workers to online retailers. There’s even skincare brand than Divya, which was able to make sense of social media conversation about people complaining about staining issues with their past products and Livia then develop an entirely new stainless product as a result of learning that there’s Edmunds. com The online automotive resource which is saving millions on their search advertising by noticing trends in buyer behavior and as they change they quickly react with changes to their ad campaigns. So the phrase dark data might be losing meeting as we all try to make sense of it but dark data is a real problem and solving it can be the difference between meaningful digital transformation and stagnation in the face of constant change

Katie, I put this to you. I think there’s a single barrier preventing a lot of executives from using their data well and I actually don’t think it’s technology. I actually don’t think it’s any kind of like algorithm. They need a team. They need to hire. I think it’s their own pride.

Nobody wants to be wrong. You know, it’s really hard to point out your own flaws and so as you were sort of you know mentioning that my first thought was, yeah, it’s ego because you know they don’t want someone else to point out to bring in front of them, hey, here’s all the things that you’re doing wrong because he didn’t know it’s a business at the end of the day it’s still taken personally like well those are decisions that I made.

So I’m doing it wrong, and especially you know the higher up or the longer they’ve been doing it. They don’t want to be told that they’re doing it wrong or, you know, they’re just like, well, I just know because I’ve been doing it and they become sort of an N of one that was one of the my favorite phrases when I was doing clinical trials or, you know, as a product manager was you can’t be an no one. You can’t tell the customer what they need. You have to ask them what they need. And I think that, you know, to your point, it is pride. It is ego. This is all about challenging assumptions we assume this works, we assume the answer is Y or Z and we can’t assume we have to actually put it to the test.

So I go to the Charlie Munger quote Charlie Munger the chairman of Berkshire Hathaway, where he says, I’d rather be vaguely right then precisely wrong and what I took away from that quote is he’s somebody who he’d rather be constantly course correcting and not have the answer in theory you know some best practice or some rule that he cleans do then actually have one of those theories or best practices find out that it’s yes it’s precise, but it’s wrong so he’s like not somebody who supposes he has the answer. He doesn’t make assumptions. He’d rather constantly say I have a hypothesis, you know, like your example of working in clinical trials in your career. It’s like you have a hypothesis, you’re out to prove or disprove it, it doesn’t matter. What matters is that you constantly course correct and this zigzag line, but you’re never going to have the answers. The important thing is to know how to find them. And I think that’s, to me, the power of what brain trust insights is trying to do.

Yeah, exactly. And, you know, there’s a reason why Berkshire Hathaway is so wildly successful is you know if they adopt that type of mentality then yeah, you’re always going to be course correcting along the way. It’s also sort of in a related way. One of the reasons why I’m such a big fan of agile methodology versus waterfall because it’s that constant iteration and course correction versus just marching in a straight line until you get to an answer, finding out the answers wrong that haven’t start all over again, you know, with agile, you can constantly be changing and correcting and iterative evolving. I think of the the evolution of like the human species a lot as it relates to like behavior in marketing and business overall it’s weird but you know we used to have to like club things to death. And then we could get more precise with the way we hunted and killed our dinner with bow and arrows, etc. etc. And we got more precise. We have more precision in what we were able to do.

And I think we’re living through that now where I think in the older ways of leading a business, you would have executives who had all the knowledge, like they they profess to know because they’d seen it and you know the more experienced they had the more they sort of quote unquote knew or had the answers. Now, I think we’re living through an era where it’s amazing. We can learn so much about what’s actually going on and get out of theory and start looking at reality through the data that we have through qualitative feedback to and actually get to know the customer and understand that better than our competition and therefore when and the executives that likes to cling to theory and say, Well, I know that you know I just trust my instincts. I know I have the answer or this is how we’ve always done it. That’s like a caveman mentality. It’s like that that’s an outmoded way of relying on leaders that kind of leader is no longer needed in business, they have to adapt and evolve to understand it’s, you know, I’d rather be vaguely right then precisely wrong.

I love a good natural selection analogy. But no, I mean, that’s exactly it. You know, people companies leaders who aren’t willing to take a look at what’s already happened or what’s happening around them and evolve. It’s that evolution that you can’t just look at what’s happening. You actually have to do something about it, you know. And so in back to your conversation with Scott Monty about digital transformation, it’s the exact same thing change is hard. Change is hard at any level, you know, small changes are difficult like I know I probably shouldn’t drink as much coffee as I do. And you know what, I can have that internal debate with myself about the and cons and the health benefits, but actually make that change is a difficult thing because we’re creatures of habit.

My dad is a chocoholic who likes to site every new study that comes up that says actually dark trots that’s really good for you. If it’s this percent chocolate so he just, you know, keeps buying chocolate, you know, like we try to and that’s another another point in that brought that that story up with a purpose here which is I love chocolate. Now the real reason is you know my dad likes to make the data work in his favor and you talk a lot about data integrity. So yeah, the first red flag you threw up was Don’t be an N of one but the other is once you get out of the end of one and actually look at all the ends. In other words, your data. People can still massage that and try to make it. Say what you want it to say as a leader or a team and I imagine that you run into that conflict a lot, you know, how do you talk more about that well.

And I think that the anecdote about your dad is a really great example of, you know, I know that the behavior that I have maybe not be so great, but I’m going to make the data work for me so that I can justify my actions and if you take that into the business sense. You know, we have seen a lot of that were, you know, we can take a look at all of the data both qualitative and quantitative and say these are the things that you need to pivot on these are the actions you need to take and we’ll still get that push back have no I don’t think so. We’re not going to do that or we’re not going to prioritize that or, you know, well, I was just reading this article written by some random guy and he said that it was okay if we kept you know bleeding money into out that aren’t performing. So we’re just going to keep doing that because that’s what that guy over there says, and that’s what we’re comfortable doing because that’s how we’ve always done it.

Yeah. The Guruiism in the business world is getting a little out of hand. We need this isn’t this isn’t a word at all. But we need more truth. Truth ism truth truthiness if I’m Stephen Colbert truthiness

I agree, you know, and I think that is.Well, again, the truth is hard and sometimes you know the truth isn’t what we want it to be. And I think that that’s one of the things that I love about data so much is that it’s just going to tell you what it is you can’t manipulate it to be something that it isn’t. It’s very black and white it very much bits into its box and then you  have to do what the data is telling you to me that’s strange, but a little bit comforting of, like, Okay, well, I know what it’s saying so I know what it is. I need to do. So consider what we’re really telling people in this episode, consider that, you know, executive used to be there because they are right or they know in some absolute sense and consider now that the shift is to care less about being right and more about getting it right and a huge part of that is unlocking or illuminating this dark data. What are some questions that I should be asking myself as an executive or what are some approaches are mentality shifts that I need. Obviously, we’re not prescribing as a list of tactics.

Because every situation is different, you know, but get me started down the right path, given all that you seek at. Yeah. So I think one of the first things you need to ask yourself is, do I have all the information to make an informed decision. I mean, that seems pretty basic, and you would assume that a lot of people are doing that but you know business can be really fast paced and people are making snap decisions.

One of the things that helped me get more disciplined was having to read a business case for every new feature that someone wanted to add and so I had to go through that process of collecting information which I’ll be honest, it was a pain in the I hated the extra paperwork, but it helped me make sure that I was asking all the right questions what aren’t we doing what aren’t I looking at, you know, have I talked to people from all the different disciplines and my company to get all of their perspective. So I have a full 360 view. And then, you know, are the decisions. I’m making aligning with the business goals and if I get nothing right if I’m if I’m just there’s all these big decisions you know if I’m only.

Making one mentality shift. What’s your advice to me like what you know what is that one thing that if I if I just start there. The rest gets a lot easier. You know, I think in terms of the way you think about it, you know, question everything, especially the higher up that you are in an organization you expect that the information that people are feeding to you is complete. If you just start asking, Is there anything else do we have all the information then people will start to say, Oh, I don’t know, maybe there is more than to the story that I could tell you, and then you’re just sort of doing a little bit more of your own due diligence on behalf of the company and then don’t be afraid to highlight something that you’re not doing right because that’s the only way that you’re going to get better back to that Berkshire Hathaway example you know they’re constantly pivoting based on information that they’re getting and they’re wildly successful. So there’s a reason for that. Don’t be afraid to point out the things that you’re not doing well.

Dark data is becoming a real buzz word, I get it.

But the benefits of shining a light on your dark data are undeniable. So ask yourself, do you see or do you observe if Seeing is believing than observing truly understanding what goes on leads to transformative action in the digital age. It all starts with curiosity and the willingness to test every assumption, you have to get it right, rather than be right Sherlock Holmes may have called that the distinction, but we just call that the difference.

The difference is the podcast from brain+trust partners, an executive consultancy that helps busy leaders make sense of customer data and customer behavior in the digital age. Learn more at braintrust.partners

This episode was written hosted and produced by me, Jay Acunzo.Thank you so much for listening to the show, we’ll talk to you again in two weeks on another episode of the difference.


Get unique data, analysis, and perspectives on analytics, insights, machine learning, marketing, and AI in the weekly Trust Insights newsletter, Data in the Headlights. Subscribe now for free; new issues every Wednesday!

Click here to subscribe now »

Want to learn more about data, analytics, and insights? Subscribe to In-Ear Insights, the Trust Insights podcast, with new 10-minute or less episodes every week.

Pin It on Pinterest

Share This