Introduction
This week I’m joined by Dave McCrory who is the CEO of Data Gravitas.
Dave popularized the term data gravity a decade ago. He starts off by defining what data gravity is and why it is important to consider. Dave talks about his book and the research he is doing around data gravity. He talks about the behavior economics of data and how it is widely misunderstood. Dave gets provocative and discusses the hand waving that happens around data including data lakes and value of data.
Speaker Profiles
Dave McCrory Twitter: https://twitter.com/mccrory
Dave McCrory LinkedIn: https://www.linkedin.com/in/davemccrory/
Data Gravitas: https://datagravitas.com
Audio File
Featured Hashtags
#CIO #Leadership #Data #DataGravity #DigitalTransformation #CxOitk
Podcast Transcript
Tim Crawford 0:01
Companies are looking for new ways to transform their business. Technology plays a critical role in this transformation. Speed and innovation in both technology and thinking are key to this shift. Hello, and welcome to the Cxo in the Know podcast, where I take a provocative but pragmatic look at the intersection of business and technology through the lens of leading CxO executives. I’m your host Tim Crawford, a CIO and strategic advisor at Avoa. This week, I’m joined by Dave McCrory, who is the CEO of Data Gravitas. Dave popularized the term data gravity a decade ago. He starts off by defining what data gravity is and why it’s important to consider. Dave talks about his book and the research he’s doing around data gravity. He talks about the behavioral economics of data and how it is widely misunderstood. Dave gets provocative and discusses the hand waving that happens around data, including data lakes and the value of data. Dave McCrory, hey, welcome to the program.
Dave McCrory 1:07
Thanks, Tim. Glad to be here.
Tim Crawford 1:09
So, to get us started, why don’t we start by having folks understand a little bit about who Dave McCrory is, and at a high level, what data gravity is
Dave McCrory 1:21
too. So, I’m Dave McCrory. I have been in the industry for over 20 years and popularized the term data gravity about a decade ago. That was brought during my time when I was working at the hyperscale compute division at Dell, the DCS. If if you’re familiar with that, I’ve been the CTO of Basho. I was the SVP of Engineering for Warner Music Group. I was most recently the VP of Engineering for Machine Learning for Industrial Internet of Things at GE Digital, and really have been studying data for a decade now, and specifically the effects of data gravity, and data gravity is something that you often hear about in the tech sector, but it’s not always explained in the simplest of terms. The way I look at data gravity in its simplest form is you have some amount of data, and it has an attractive force on the things that create and consume data, which would be primarily applications and services. The reason it’s attractive is the closer you are to the data, the faster the access that you have to that data, and by faster access I mean lower latency over a network, and/or higher bandwidth. So you can either get more data at a time, either put in or taken out, and/or you can get to that data more quickly. So that’s why data has an attractive force at its at its core.
Tim Crawford 3:02
Got it. And you know we’re going to delve a little further into data gravity and kind of the ins and outs and why people should really be thinking about this. There are two other pieces that that I wanted to touch on here. Is you’re also writing a book. Why don’t you talk about that and then also some of the research that you’ve done around data gravity.
Dave McCrory 3:24
Sure, I’m writing a book that really covers the vast majority of my research over the decade, which includes a lot about data gravity. It also talks about the behavior of data and the what I like to term the behavioral economics of data. Data is very misunderstood, in my view. Really, from executives all the way down to quite a few technical people, including data scientists and such. I think everyone has the broad concept of data, but they don’t really understand the nuances of why it works the way it does, and what’s missing. So the book addresses that. The book explains what’s being missed around data, how to value data, and and really what what can be done to resolve that, and how businesses can really maximize the value of their data.
Tim Crawford 4:21
When you talk about data, I’m always fascinated about how you look at it because I don’t think it’s necessarily the common way that that people talk about or that people consider how they use data. And you’ve said that there’s a lot of hand waving around data, so let’s get a little provocative here for a minute. What do you mean by companies that are doing that hand waving around data, what does that mean?
Dave McCrory 4:46
I think specifically, it’s it’s the idea of throwing every bit of data that you have into a data lake and leaving it there in hopes that one day it may be worth something. And you’re not getting any value out of it now, or you’re getting some value, but you don’t know really what the overall value you’re getting is. I think that’s a lot of hand waving. I think understanding the value of of data is pure hand waving right now. There’s no established accounting practice for valuing data. You have people that say that data should be valued as an asset, and it should be on the books as an asset. But there’s no defined method of how you value this specific asset. There’s a term used when applied to things that are not depleted when they are consumed. It’s called non-rivalrous in economic terms. Rivalrous things are things that you use up, and you know if you use up oil, once it’s used, you don’t get it back again. If you burn it in your car or whatever, data you can use again and again and again, and so it’s non-rivalrous, and so it becomes more difficult to value a non-rivalrous good. That’s a small example of a lot of the hand waving that’s occurring today around data, and no one really seems to want to address. Is
Tim Crawford 6:05
this more of an issue where people aren’t thinking about the value of the data, or is it that they’re just treating all data the same, or some maybe something in between, or a combination of both?
Dave McCrory 6:19
I think it’s both. I think there’s a treating all data is the same, which it’s not. There are incredible differences between different types and uses of data. The same data can be used for different purposes and and have vast differences in both its behavior and its value. At the same time, I would say there’s an incredible misunderstanding around data, and people based on that aren’t effectively using it, nor are they really doing the effective things with it. It’s more or less a lot of guesswork and assumptions that data is used in a single modality.
Tim Crawford 6:58
You mentioned that data gets thrown into these big data pots, you know, the data lake, and some have even said it turns into a data swamp because people are just throwing more and more data into it, not necessarily thinking about what that is. It’s almost the the exact opposite end of the spectrum from when we talked about data warehouses, right? Data data is incredibly structured, and then you’ve got the data lake, which might be a completely unstructured component. But let’s just kind of maybe back away from that a little bit and talk about big data. You know, it’s been a really big thing, but is it really? What’s your take on that?
Dave McCrory 7:41
My take is, you know, big data started out as having lots of promise. Everyone believed that this was the next big thing. Businesses invested a fortune into all of these data storage capabilities, and you’re right. It was actually, I would say, before the data warehouse, it was simply databases, and what we had was this progression from data storage and technology, computer equipment being incredibly scarce and expensive, to being incredibly inexpensive and ubiquitous. And because of that, we saw all of these changes occur. We went from having just databases, we went to data warehouses, and we evolved to data lakes and the cloud. And data continues to grow, but our effective use of data is really going in some in some different directions. And I think the problem of the data swamp that we were talking about is one of them the idea that if you just throw everything together, it’s going to magically become more valuable, or it’s going to become easier to use just because it’s all clumped together. I don’t think that’s true. I think it’s thinking about the data, how it’s used, why it’s used, and a lot of people are talking about machine learning and artificial intelligence is that will revolutionize big data and such. Well, I think it can help, but all it does is simply look at patterns and look for patterns. And you may uncover some patterns, but if you really don’t know where to guide it or where to look, you’re not going to just magically discover these patterns and things, and then all of a sudden get all of this incredible value out of this data, which, by the way, still costs you money to store and keep online, no matter where you’re keeping it.
Tim Crawford 9:33
But Dave, this sounds like you can’t win for losing here. On one side, you don’t keep data because you’re not sure that it’s valuable, and so you throw it away. On the other side, you throw it all into a data lake with the hope or promise that maybe down the road it will become valuable. But of course, there’s a cost to doing that. Seems like there’s a cost and a risk to both equations, both options. So what do you do?
Dave McCrory 10:02
That’s where you have to understand what the value of the data that you’re storing and what the potential value is versus the overall potential risks. And I would compare it to in our lives we don’t keep every single thing that we come across that we think could eventually have some value. We eventually decide what have I used and what have I gotten value of and what haven’t I? And we toss things. We toss them as they break, as they wear out, as they have other problems. We get rid of them, or we just have too much stuff. Otherwise, you’re a hoarder, and that brings a whole bunch of problems because now you have you don’t have enough space? You’re dragging all these things around. You can’t keep track of or find your things, and by the way, those things-the value of those things depletes because of that. Because you simply can’t keep track of it all. I think the same thing exists. There is a risk you could lose out on something that has potential by getting rid of some of your data. The question is: Are those log files from six years ago about your system really all that valuable? I’m going to bet they’re not. I’m going to bet the the risk and the value is probably low. Is the formula to Coke to Coca Cola company valuable to keep around in perpetuity, absolutely it is. But you need a structure and a model to reason about these things, and there is no model available today. It’s something I’ve spent a lot of time researching, and there are no models to understand how how data works to be able to to begin this process.
Tim Crawford 11:39
The other piece that that you’re also talking to is you need a construct to be able to apply that value equation, and that’s based on certain decisions or assumptions that exist today. So, for example, today you might say this data is those log files. For example, as you mentioned, they might not be as valuable today, but maybe they do change in value down the road. Or the converse could be true: they’re valuable today, but they’re going to be less valuable down the road. Right? As as time goes on, how do you start to rationalize that over time? Because it’s not just about a point in time, it’s about over time.
Dave McCrory 12:24
That’s right, and and that’s one of the keys is that it is over time, and it’s understanding what the time value curve of your data is in different time frames, and that is something that that is incredibly misunderstood. People believe that data has a single value curve that it follows, and that’s it. And that’s simply not correct. Data has many time value curves depending on what that data is and how many contexts exist around that data. And that’s not a model that’s used today. That’s a model that that I developed. But from what I’ve seen, it actually works in applying these different things to data to understand how the value of data can be applied. Without that, I think it’s we’re back to guesswork. We’re all just guessing how something might or might not work out in the future. You can certainly get an idea of the present, but the future becomes an impossible thing to do.
Tim Crawford 13:28
But those models that you’re referring to, those also change over time too, don’t they? The value models. So as you value data and you put in that construct, that changes over time.
Dave McCrory 13:41
Not the way you value the data, but the variables around valuing the data can change. That’s certainly possible. You could discover a new context for its use, and therefore, that might add additional value to the data. But overall, it’s still going to fit into that same model. It’s just a matter of if you discover a new context or if a situation around around the overall model changes, where all of a sudden something becomes more valuable due to an outside influence, that’s more of a what’s the probability of that happening?
Tim Crawford 14:13
Sure,
Dave McCrory 14:13
but that’s business in general. Look at the current situation of that we’re all in from working from home and such. Very few people and very few businesses had a plan in place in advance to handle all of the effects.
Tim Crawford 14:26
Yeah, there’s this interesting relationship, right, between data and software, and we’ve all heard the the phrase “software is eating the world. I know you have some very interesting perspectives on this is software eating the world.
Dave McCrory 14:44
I think software was eating the world. I think software was eating the world a decade ago, as Mark Andreessen said. I think that’s changed today. I think data is eating the world now. I think data is what is driving the world. And software is really, in my view, taken more of a back seat. If you are in control of the data, if you are generating the data, or if you understand and are able to leverage the data, you’re in the driver’s seat, and everything is beginning to emit data. By everything, I mean from watches and and wearables to home appliances to vehicles, everything is now emitting data, and we’re outfitting more things to emit data. So I see that as being the great enabler. It’s not software. Software, as I said, was and and and it certainly was. If you look at if you even look at the iPhone and you know the App Store and the Android ecosystem and Google Play and all of that, software was eating the world in those aspects. But now it’s data, and I think it’s going to be data for quite a while. The two go hand in hand, but without data, software is useless. Sure, you can’t do anything with it.
Tim Crawford 16:04
But I could see an executive listening to this and saying, “Hold on a sec, Dave. 2030, 4050, years ago, we were using data to make business decisions, right? So, is the issue that we now have more sources of data that provide greater degrees of granularity and insights for our business decisions, or is it something else?
Dave McCrory 16:32
I think it’s definitely, as you said, more data because everything’s generating data, but we also have the capacity to do more with that data, and we have that because processing power has exploded. The open source movement and the availability of all of these algorithms, machine learning has finally begun to rear its head as something that can be useful for things. The internet and the connectivity that’s happened has freed the flow of information, and we see that, by the way, continuing to move forward. If you look at 5g and low Earth orbit satellites, all of those things do nothing but continue to amplify that spread of data. So you have more data moving faster with better algorithms. So a lot has changed.
Tim Crawford 17:22
Got it. Should businesses be concerned about data? I mean, if I look at at this from the other perspective for a second, how should I be concerned about data? I mean, there’s you’re talking about a lot of great things that come from data, but are there concerns that I should be thinking about too?
Dave McCrory 17:41
Well, I would say definitely. I think there are a lot of things that are overlooked until it’s too late. If you look at all of the data breaches that have been happening, if you think about, in some cases, companies are storing data for an incredibly long time, yet they’re not updating security controls, compliance, governance, and other things over this data. So really, what they’re doing is they’re storing this data while they’re increasing their risk and exposure. And not all data will help everyone in all situations. It can be harmful. We’ve seen that I think quite a bit, especially in the press. And I think it’s often neglected that the the value equation also applies to the ability to end up with losses due to legal ramifications and violations of the law. If you look at all of the different possibilities, even if they’re unintentional, there’s quite a bit of quite a bit of risk, and I think because people aren’t and haven’t been able to really model all of these risks because they don’t have a way of applying and looking at what the cost could be, they’re actually increasing their exposure because of that.
Tim Crawford 18:58
You know, risk is is such a critical component at the board level and the C-suite level, and we hear a lot about the data breaches in the public space. But as you mentioned, you know, regulatory compliance and privacy are are a key aspect and trust in there. That seems like just a complicated space to navigate.
Dave McCrory 19:17
I think it is complicated. I think with the right training and tools, and with a focus around data through this new lens, you can at least make progress, and you can begin to reason about riskier things-things that would fall under greater risk, or that you would be more concerned about-and then really begin to start systematically wrapping your hands around compliance, governance, risk exposure, and and by the way, how do you ensure yourself to minimize the risk past even you know whatever the best firewalls, security, or processes you can put in place?
Tim Crawford 19:54
Sure. So I think about the risk piece. So got that, and I think about the. Challenges that come from big data and data lakes. Where would you suggest that businesses really need to get value from the data they have? I mean, where where do you start in that conversation today? If you’re maybe you’ve tested some or worked with some of these technologies, but at a business level, I’m looking to get the best value from my data. What do I do?
Dave McCrory 20:26
Well, and I think that’s that’s really the heart of all of this. Is I think the quote-unquote leading companies are the ones that have created created feedback loops where their data actually not only gets them customers, but gets them more data that they can then use and get value out of, and that’s more of what you would see out of a Facebook or or Google or something like that. And I think that’s what most businesses would would love to achieve those margins and levels of efficiency. So, how do you go down that path and get the value from your data. I think it’s first understanding the data you have and the nature of how it’s being used today, not only internally, but how it’s used externally. And that’s not focused on by virtually any company. There’s a very small list of companies that focus in that way, especially if you look at more more established companies. They’re not looking that way. They’re not going and investigating how customers are using and consuming their data and what data might their customers be making available back. They’re not looking at how they could leverage that. What are the different ways? And really exploring and going down that path is really a way to begin to understand how you can add value. And as I said, there’s a way of doing this. There’s a systematic way of doing this that you can go in and you can build this model. And then once you have this model, you can apply it. And will it be the perfect model, and it solves all problems? Absolutely not. But will it get you leaps and bounds ahead of where you are today? Absolutely. It gives you a way to reason about how all of these things work and how to actually get that value back out. Commonly, it might be combining or linking your data to other things. It might be looking at deriving data or even selling some data. You might be able to make additional money by making your data available to another party, a customer, something else, while not really increasing the risk.
Tim Crawford 22:36
Interesting, interesting. So as we we kind of get toward the end of our discussion, I want to bring this full circle back to data gravity. So we’ve talked about risk, we’ve talked about the opportunities, we’ve talked about the data components. As an executive, why should I care about data gravity? Wouldn’t that be something that would be further into the organization or a discussion further into the organization? Why is the C-suite looking at or should look at data gravity.
Dave McCrory 23:04
The C-suite should be looking at data gravity because it can provide guidance on where they should be pointing their business in the near and long-term future. I think data gravity is a great predictor as well as limiter on what’s happening within the flow of your business, your business communicates using data. All businesses do today, and data gravity is a great way of looking at how that data is flowing, both to and from customers as well as partners and internally. And if you look at it in that way, understanding the kind of centers of data gravity can explain where you should be going, what you should be doing, why you should be doing it. If you have a reasonable view into this, you can begin to look at how your business operates in a new way. And any time someone presents you with a way of viewing your business differently than you have up to this point, it’s an opportunity to learn. And I think all C-suite executives having the opportunity to gain a new insight into their business without having to be a technology expert should be and would be interested in this.
Tim Crawford 24:21
Are there things that that folks are doing that prevent that from happening?
Dave McCrory 24:26
I think a lot of people view it as it’s just data, or that’s just the technology stuff, or the infrastructure, and it’s viewed in that way, and it’s not viewed as everything the business is doing is flowing in this unified format, and if we have a way to unlock looking at this unified format of how everything in our business works, we can get unique insights and make unique discoveries that we don’t know, both self-reflective, in other words, looking internally, as well as looking externally at what’s happening. And I don’t think a lot of companies do that. I think they just look at it as we do these things, and yeah, there’s data because I store things on my hard drive and I do those things. But I think that’s the level that data’s looked at, or it’s well, look at the financials. This is the data we’re looking at, but they’re not looking at at data from that higher level than I’m describing.
Tim Crawford 25:22
That’s awesome. We’re gonna have to leave it right there, Dave McCrory. Thanks for joining the program today.
Dave McCrory 25:26
Thank you.
Tim Crawford 25:29
For more information on the Cxo in the Know podcast, visit us online@cxointheknow.com You can also find us on Apple Podcasts or wherever you listen to your podcast, please subscribe and thank you for listening.
Discover more from AVOA
Subscribe to get the latest posts sent to your email.