Simplify for Success - Conversation with Aaron Crews
Aaron shared his thoughts on data analytics and governance based on his vast experience in eDiscovery, IG, AI, Big Data, and Analytics. As the first Chief Data Analytics Officer at Littler, he recognizes the importance of curated #data and discussed how effective information governance leads to better data analytics, resulting in better decision-making and more insights. Because actions speak louder than words, he believes demonstrating the value of data as a strategic asset can lead to better investment and scaling.
Listen to the podcast here:
*Views and opinions expressed by guests do not necessarily reflect the view of Meru Data.*
Hello everyone, welcome to our podcast around simplifying for success. Simplification requires discipline and clarity of thought. This is not often easy in today's rapid paced work environment. We've invited a few colleagues in data and information governance space to share their strategies and approaches for simplification. Today we'll be talking with Aaron Cruz.
Aaron is Littler’s first chief data analytics officer, led the firm state analytics practice and big data strategy. His professional experience includes time spent as an associate and shareholder at Littler. A stint as a senior associate general counsel, responsible for E discovery at Walmart and time as General Counsel and Vice President of Strategy and an Artificial Intelligence startup.
He has lectured and published extensively on topics of E Discovery, Information governance, big data analytics and artificial intelligence, and is particularly well known for bringing cutting edge solutions to the most difficult problems faced by clients. Hi Aaron, welcome to the show.
Thanks for having me.
Why do you think data ownership is critical?
You know that is probably because of what I do right. So, the thing about analytics is analytics is only doable and it's really only useful if you have good data and so the data ownership issue, right? And when I use that phrase, I use that phrase really, you know, writ large.
But the data ownership issues or the things that stifle or prevent you from getting an analytics program off the ground, and that's whether internal facing analytics or product related analytics, you know you don't have either. You don't have access to the data. That's hugely problematic for obvious reasons, but maybe equally problematic is the situation where the data is there, right? And theoretically, there are consents around the kinds of uses that you want to put it to, but the data is spread here, there and everywhere, right? It's not been consolidated and tagged; you don't have single source of truth.
Those kinds of problems make the ability to do analytics really difficult and can really undermine a program or undermine a product, and then you know taking it a step beyond that. You have this kind of fast-moving world of machine learning. And in that machine learning space, now I'm a big believer that if you don't do analytics well, you're not ready for machine learning. So, you know, if you've got those data problems on the analytics side forget about diving into machine learning anytime soon, because the kinds of problems that stifle analytics programs will kill a machine learning initiative.
So, for me it's really kind of a mercenary need that you know good data, ownership, good data, governance, understanding what we've got, where it is, what it is, how did we get it? What kind of consent did we get around it? All of those are limiting parameters on your ability to do really productive things through the analysis of data. You know whether that's being done by kind of data analysts really getting and crunching it, or whether you're doing it via algorithms or some hybrid of those.
So, you mentioned a lot of things, and so maybe I'll kind of start with a few things, right?
So, you talked a little bit about knowing where your data is and if you don't know where it is or that it exists or that it's a single source of truth you can't really do anything. So, we've had the concept of data stewardship for a very, very long time now, but it hasn't worked at all. Why do you think that has never worked? And what are some of the problems with the data stewardship program as it is today?
Uh, you know my answer is always follow the money. If there's not a business value to true stewardship, then people aren't going to put the resources there to do it. They're going to do whatever they have to in order to meet, you know, black and white obligations, but they're not going to put in the resources that are really necessary to be able to, you know, to do data stewardship in the way that would be most effective, you know.
And that's why in a lot of organizations you have kind of to the extent that you have data stewardship going on. Think about this a little bit right and a lot of organizations you have certain kinds of data that are sort of being stewarded if you will, right, but a lot of that stuff is actually the great unwashed masses of data. It's just kind of collected in places you know it's not being carefully thought of, it's not being cultivated. It's not being acculturated if that makes any sense you know or if there's a better term, curated would be a better term, right? It's not really being curated because there's not a recognition that there's value.
You have folks in the business who will look at it and go, yeah, yeah, you know the nerds want to do whatever with data. And I think that's fundamentally how you drive the shift from kind of the old view of data stewardship as an unwanted obligation to the concept of data stewardship as an asset that helps facilitate improving the business through better productization right?
Do you think this is more of a chicken and egg problem like in the sense data analytics really struggles because you don't understand data, or the data is not well maintained. But the data is not well maintained because data analytics doesn't recognize it as a necessary need for them to kind of move the needle or you know, enhance the bottom line.
I think it's the second answer. I think it's the latter, not the former, but with a tweak I think the folks who do data analytics, they understand the value, right? They're the ones in there banging the drum saying hey, we got to do this and here are all the reasons. Most of the time there’s a gap in the businesses’ understanding of the value of data analytics, right?
I think a lot of times, unless it's a company where you know analytics, or you know the development deployment of algorithms is really core to the business. I think a lot of times, a business thinks of data analytics in a very transactional kind of way, right? They're looking at internal data telling me about this thing about my business acts, and I think they maybe don't think in a broader sense of what data analytics could do. And how really kind of pushing that envelope could really help develop and expand the business You know, like giving you a very like brass-tacks real-world example, right?
Some of the stuff I've been working on for the last several years is taking data, structured data, and layering that structured data together in order to get a very clear picture of what people are doing and when based on the allegations in a lawsuit. So, like a complaint is that a group of people worked tons of hours of overtime that they never got paid for. They never took any lunch breaks, and they never got any rest breaks. Invariably it's like there are thousands of people that are implicated and the plaintiffs in the case want you know, $100 million in order to make the case go away.
You know one of the conversations that we end up having with those clients is there's a way to do this that you know partners have historically done this, where they're going to depose everybody, they're going to ask for emails, they're going to do all this stuff that's kind of a waste of time.
And the better way to do this out of the box is to think about what systems those people are touching, and they have to be touching to do their job. And then to use that log data to really build a picture of what a person's day looks like. And you can do that for you know every day during the class period for every single member class, if you can get to the data.
It's a really frustrating conversation to be able to say look, we could solve this question for you. We could tell you exactly what the liability is, and the damages are in this case. We could do that like the 1st 90 case if you had the data, but because you guys haven't been keeping it or because it's spread all over the place and it's in seven different formats, you know that it becomes impossible or exceptionally difficult, right? And then we're having conversations about other kinds of data that can stand in for the stuff that we really wanted and all of that.
Increasingly the Clients that we work with, that are experiencing this, they start to ask questions like, well, what data should we you know, should we be keeping how do we keep it? that use case creates a Push for the kind of data stewardship that you know we that we're talking about in this conversation.
But until you surface that for them as a capability and as a value proposition, I think it never occurs to people that they should be paying attention on the front end to data. And then I think the other problem quite honestly, is there's so much right especially if you're a large organization, there are so many systems that are coming online.
And when people are thinking about what they're going to do with that data a lot of times they're thinking about information security. They're thinking about their regulatory requirements, right? But what they're not often thinking about is, are there downstream business uses for this data, and if there are, what are they?
I mean, if you really want to shift the stewardship conversation, there's got to be kind of a more holistic conversation with those of us who do data with the business, that is a very business driven conversation that says data is a resource and like any resource, it's only valuable if you husband it a little bit.
And so, we should make a concerted effort to husband data in a way that moves the business forward and allows us to do things in the future that we don't recognize today would be valuable. Right, but you know they will pop up.
Here are some examples we've seen in the last couple of years. We know there are more things coming like this. Let's make this part of the program, and so I think you know it's a long-winded way of saying essentially, if you can wrap yourself in the business and then the business is success, I think data stewardship is likely to happen as long as Data stewardship continues to sort of be thought of as an unwanted obligation, I think it's a really difficult thing to get people to do well.
So, you mentioned something very interesting, right? So, you talked about the other side of the equation, which is you know security, privacy, risk and of course you know records management or data minimization efforts.
I mean, when you walk into a larger organization, you know you're talking about 700, 800 thousand systems, right? Very easily that's not you know out of the ordinary. So, those who are talking about regulatory requirements think in terms of, well, you need to keep this data, because this law says so.
But then if you sort of look at what is your IP or what is your asset for the organization, that's going to generate maybe more value, more insights about your employees, more insights into your operations, more insights about customers, whatever you’re spending time on. If you look at the security classification, if you look at the RIM classification it tends to go into admin, right?
So, it gets into like a or it's a non-existing bucket. Well, which you can call it admin, but it's not
Yeah, yeah. It's a bucket called other as I used to say right when I was in-house, I would always refer to the bucket called other. Which was everything, like if you looked at it on balance right? We probably had a tiny fraction of what we were doing that went into buckets that were actually validly titled right, and everything else was kind. Of like Oh yeah, it's other.
People say 70% of the data is raw, delete it. But if you have put all of that in the other, how was anyone going to delete it? I mean that's the most important pieces of information, and if you can't give the confidence that you've got that figured out, you know obviously 70% of the data, nobody is going to pull the trigger.
So how do you build that bridge between different parts of the organization, privacy, security in information governance and then IT and of course your data analytics Team and the businesses.
I mean, that's a big question. We could spend a while on this one, I will invariably get longwinded here. So please cut me off and interrupt me as you deem appropriate.
But I guess the way I think about this is building a really successful program actually creates those bridges, that's a cross functional team, right? So maybe that's part of the first problem. The first problem is you've got you just named 6 organizations that all have responsibility and ownership interest you know authority over however you want to characterize that right their jobs, right? They all have some level of mandate around this data, and those mandates do not necessarily align right
So ITs mandate is pipes and plumbing right? bring in these systems often kind of the new shiny things, but not always, because maybe the business is asking for it, or maybe they just think it's cool or whatever, right? So, you have IT bringing in systems and kind of knitting them together and making them go, you have privacy that's worried about kind of its sphere, right
You have information governance which may overlap with privacy and may not because they've got at least conceivably they've got a business mandate right around data that the company has deemed to be valuable, and so that's going to be actively curated. And that may just be like, you know, your CRM or whatever, but it might be something far larger than that. And then you know you've got the lawyers who are involved. So OK, that's there's that group.
Then you've got data analytics in business, right? And the business, I think unless they see you know the value that that data analysis can bring to the table. I think a lot of times, except where it's obvious the business is kind of like I don't care. Right, uh, that's kind of not in my on my dashboard so to speak.
And then the data analytics folks like if you go and ask a lot of us would say we'll keep everything. I keep everything. Because there is a tremendous amount of value that you don't even understand yet, that's in that data.
And if I start to mind that, I can tell you all kinds of things about all kinds of things. But I don't know what data is valuable until I know what questions you would like to answer, because the thing about data analysis. Data analysis only works well when it is backed by or when it's done in order to answer a really good question right. Really good, very narrow questions help you figure out what data you need to look at and how do you need to look at it in order to kind of elucidate the answer to the question.
You know, if you put the data analytics folks on one side of the number line and you put kind of the regulatory privacy legal folks on the other. You've got kind of a big chasm between those folks, right? And some version of that group you know, data analytics is like we keep it. This is essentially the new oil.
Nobody throws away oil like when people were making kerosene. They threw away oil all the time because it was a byproduct until people realized it was really, really valuable, right I don't know where I know that, but I do and then I think there are a lot of people on the data science, data analytics mill side of the house who would argue that people who are dumping the rot are essentially they're living in a kerosene world and they're throwing out oil.
And then you know you have kind legal and sort of regulatory and compliance folks who are saying keeping all of that data creates risk, right? And because we don't understand what it is and because we don't have an application for it today, that's potentially valuable and makes money, that risk may not be reasonable to keep right.
And then you've got the IT folks in the middle, you know there's the systems they're responsible for, and they're bringing in. Those are generating the data. At the same time, maybe under budget pressure or whatever to shrink the footprint because while storage is increasingly, you know, moving towards zero, it's not at zero, so there's a cost to that.
And I think you know all of that together plays to create like fiefdoms and little walled communities as opposed to bridges at help have that conversation. I think it's a really progressive organization that kind of puts all of those voices in a room on the regular and has them work together to come up with a plan that makes sense for the business. It says to the data analytics folks like look, we're going to have to throw out some of this stuff that you think is oil. So, let's kind of pick and choose what that is. If for no other reason than to make the legal compliance folks happy and to help IT with their budget.
I think if you were going to build those bridges, would have to do it in one or two ways. You have to put somebody in charge of those data issues and give them a mandate over all of these kinds of competing organizations.
Or alternatively, you have to bring those competing organizations together on the regular in a room and you have to say to the Leaders of those organizations, you collectively are responsible for building out a plan for how we're going to do this, how do we make sense of this? How do we decide what's valuable and then? How do we curate data going forward, right? Because you know, cleaning up the mess is always harder after the fact than trying to get a handle on things as it comes in.
So, you know if you can start to figure out, kind of that, what do you have? Where is it? Where is it coming in? What's it? Comprised of? If you can start doing all that legwork you know on the front end, then categorizing the data and pushing it into the right buckets is a much easier process than trying to sit down and triage that big bucket of other and figure out what that stuff is.
Because a lot of times it's untethered from how it came in right? You might not even be able to tell what that's not sometimes and then you know you get into the whole what do we do with that conversation.
Because if you don't know what it is I don't know how you do anything with it and you're not sure what it is, what the consents are what its value is blah blah blah, I'm also not sure how you get rid of it because you don't know what it is, how it came, in, what its value is blah blah blah.
You said again, a lot of things, so I'm trying to think which one you know, I can take this in so many different directions. So, part of it is you just mentioned this whole gap between you know the those who are enriching the data and then the others who are sort of trying to govern the data.
What about the same level of gap? I mean I do think that and now I'm kind of shifting the focus on technology. Obviously, technology plays a role. Yes, you can kind of assign all of this to one person or figure out a way to collaborate, but you need something to sort of enable that collaboration, even if it's if everybody is reporting to one guy.
So, part of it is I kind of look at technology as you have some things that are pie in the sky. where it's more of a strategy or a plan, or a or a couple of you know ideas that sort of organizes the thoughts. Or maybe it's a couple of spreadsheets or technologies that sort of, uh, take and put a plan together and then you have, you know you exactly kind of name the problem, which is classification is very easy, but you have not sort of organized the classification in ways you sort of have this forest of large amount of data, but you can't do anything with it. At least the energy that it will take to sort of just run Or at least, that's the general consensus, which is, you know, once you get can somehow take the massive amount of data and put it through some tool you can now do whatever you want with it.
But you know, oftentimes you're left with probably large amounts of data that you might have some sense, but you don't know where it came from, what's the consent? You've kind of lost the bigger picture view of it. So, what do you think in terms of do you think that is a challenge from both the data analytics as well as governance perspective in terms of marrying the two?
Yeah, for sure. I mean maybe that is the ultimate kernel of the problem. There really isn't any magic button, right? There are technologies that help. Some of them are very good, but again, now you're back to the realities of the business, right?
You know you can have the best idea in the world about how to triage a bunch of data that is kind of this deadweight we can't use it. We can't get rid of it; it's costing us to keep it. We don't know what's in it, its kind of a problem, right? But that deadweight data issue it's got a cost right and that cost is essentially the status quo.
And then if you want to do something with that right, I mean businesses are in the business of making money, so. You know part of. The way you make money is to only spend money on things that are valuable right? If fixing the status quo costs a lot more than the status quo costs. You know, even over some period of time, most organizations aren't going to make that investment, right?
That's kind of the first piece and then the other way to tackle that conversation is to is to articulate the value of curation, right? So, there's not just a cost mitigation piece of it. There is a positive value and I think sitting down and articulating that helps a lot because you know the systems that people bring in to crunch through tons and tons of data, even you know if that's possible.
And in some organizations that might be possible in a lot of organizations, it's not. Because what the systems are or where they are, or, you know their key to running the business. Right, and nobody is going to put something in there that's going to kill bandwidth and shut off all your retail locations or whatever. Like those are really kind of nonstarter, sort of proposals a lot of the time
But assuming that it's doable then and you know there are situations where it is, so assuming it’s doable. The folks who are involved have to put their business hat on and think of it as a as a business problem as a business proposal. So, our current run rate in the status quo is X. There are risks associated with that that are Y, and you know, you might not be able to quantify exactly the dollar value of that risk, but you put a ballpark on it a lot of the time.
And then you know if we are able to effectively triage this data, then what we'll have left will be really valuable for analytics and now these business pieces that you know the business is trying to advance. We can help inform all of that with well curated data that has a value that is Z. Right, and so X + y + Z is some positive number and that compared with the status quo, right? It makes sense to make that Business move.
I think setting it up like that creates the opportunity for success, but I think that a lot of the time, again, usually because there's not one person who's responsible for it and how do you cobble together sort of that business plan, right for the project? Uh, who's going to do that? Who's going to take ownership of it, Who's going to do it well?
I think that is really kind of the gap. You know the technology piece of it like. Technology is like, uh, you know. Hammers, or technology, right? Screwdrivers or technology? Uh, we always think of technology as software, but you know, software is, it's a modern hammer at the modern screwdriver at the modern whatever.
So, and I say that just to point out. That, like you know, there's tech out there. A lot of the time that can solve these kinds of problems. If you can create the environment whereby that makes business sense and the only way that I know of to create that environment is to build the kind of business plan that I was just talking about, and I think in most organizations that doesn't happen and because that doesn't happen. You know the kind of things that should happen on the data side just aren't.
So how do you scale data ownership?
To be honest, the answer is I don't know. Uh, because I don't know anybody that's done that well Again, I come back to you got to have one person who's responsible and you have to give that person resources. That team has to be aligned to the business and responsible. Or are owning and curating data in a way that moves the moves the business objective forward.
If you do that and if that enterprise starts producing results, right? I think you see a natural organic scaling because people you know businesses would put resources behind things that add value. Things that add value and things that shrink risk.
It's like growing any business, right? If I think if people who were responsive data stewards thought of stewardship as their business and they approached it from a, I'm going to add the kind of value that brings, you know that that gives me a really strong argument that you should give me more resources, right? I think if people approach it that way really, like. You know, almost like a startup mindset to you know their function. This function is my business. The job of this business is to, you know, provide valuable products in the form of data that can be used for valuable things, decrease risk and overall decrease costs associated with things like raw, right I think something like that could produce the kind of business plan that garners resources because it makes sense.
You're able to pitch it as a business to businesspeople and the thing that everybody kind of, I think in our space forgets is that the business trades dollars for risk every day. So just coming in and saying that's risky is not enough to get them to move. There's got to be an additive value that is again in the form of decreased cost and probably improved outcomes.
Because decrease cost alone again, probably not enough, right? The phrase I always use, people answer will just pay for the gas. Right, you've got a muscle car. that you really love. It doesn't matter the gas is about to be $6 a gallon out here in California, right? They're still going to pay for the gas. It doesn't matter that it gets absolutely garbage mileage, right? They're not going to buy a Prius. And the only way to get them to buy a Prius is to say essentially there's a far higher value to your Prius than there is to your muscle car. You'll get more enjoyment out of it, you know, blah blah blah.
A business version of that kind of does the same thing and does away with the, I'll just pay for the gas mentality I think, and I think if you can do that, you have a shot at scaling ownership.
I guess the other way to do it, and this is something I've been thinking about for a while, you know. I've had this concept for a while of a data trust, where there are financial incentives associated with stewardship, essentially right? That people could opt in to giving you their data or specific purposes. And you could compensate them for the uses of that data, and you could set it up in ways that are really smart. Like you know, really valuable data. You compensate them a lot more because you can get a lot more. You can charge a lot more for it.
So, you know health data is rare but not crazy rare right condition. That's probably worth a lot more than like you know my Netflix watch history, right and so I think that might also be a mechanism, but again it it's really kind of a market-based mechanism, something that you know drives a business is likely to engender scale. Everything else is likely to engender a begrudging Minimum resource investment and not much more.
Sounds good. What do you think of a data lake?
I love data lakes. I think they make all the sense in the world. But you know the thing, the thing about data lakes is you need to know what's going on in there, right? And I think I think there are kind of different views or maybe different flavors of data lakes. There are some people that say like just throw everything in there, right? And then you can allow you know your data scientists or your analysts you know you can allow the nerds to take what they want out of the lake. And use it and whatever.
I think there are some issues with that. Uh, amongst them, right? If that data is not curated in some substantive way, the analysts may see data, think it's something, and have it in fact, be something else that's not entirely abnormal in any way. So, I think data lakes are really valuable.
But I think a really functional data lake needs to be minimally curated at the least right you need to kind of, you need to know what's going into it. So that there's not superfluous or contradictory data in there. A data lake is different from a data ocean or a data dump, right? A data ocean or data dump is just where you throw everything, and we'll sort it out later. It's the other bucket that we were talking about earlier.
I think, uh, Data Lake should be a little like a golf course or like the gardens at Versailles, right? Like it's wild-ish, but it's not. The state of nature is scary. Hobbes was not wrong about that. And curated is probably the better play.
And again, you know. The upside of a curated data lake is it allows the constituencies that are worried about rot and risk and other things, how to be in that conversation and to move their priorities forward at the same time. You know what's in it and you know how it's structured. You know why it's in there, then when it's no longer valuable, you can deal with it in ways that bring the blood pressure of the compliance risk folks down, I think fairly significantly.
One last question. Would you treat Structured versus unstructured versus semi structured data differently in the sense. Would you have a different strategy for managing them or do you think that they should kind of come all under the same umbrella?
You know I spent $100,000 going to law school, so I want to give you the lawyer answer. Uh, which is it depends. In all seriousness, I think you need an omnibus structure. But I would say, generally speaking, your structured data is probably more valuable than your unstructured data over the long term.
You know people, people hoard email, email is a classic example of unstructured data. I was a part of an organization where at one point I interviewed a guy about his email and he told me that he had every email he'd ever sent or received at the company in his like 10- or 12-year tenure, right? And when you ask him why he's like, well, just in case. I need it. I mean like how many times have you gone back and looked? He's like I, you know I do that on the regular and then you ask him the money question, which is how far back do those searches usually go? and it's you know, it's the last year. The last six months. The last three months. It's not the 9, 10, 11 years preceding the current.
Structured data tends to be valuable and tell you things about the business and the people who are operating in the business and around the business when you analyze it right. And so, like the example that I gave of the wage and hour litigation is a really good example of finding value in structured data.
Unstructured data. And because it's unstructured, it tends to be far less useful. I mean emails are being written presumably I mean, setting aside all the CYA emails that people write. For the most part, it's being done to communicate things in order to get things done. Structured data usually shows the things actually being done, and so you know if you need to know the ‘Whys’ then maybe you need to dig around in the unstructured stuff.
But if you need to know the what and what is usually what matters, you can oftentimes you can figure out the ‘Whys’ from the what you know that's going to be mostly in your structured or your semi structured data. The thing is that you've got a rubric around them. You can understand them,
And also, I think the other thing about structured data is it's easier to draw a line around staleness for structured data, right? I mean if you've done. If you do any kind of data analysis over time you start to realize pretty quickly where the area of diminishing returns begins, how far back you know the area of diminishing returns begins, in looking at data.
And you can use that you know, as a guidepost for what you're keeping, because, generally speaking, right? My thinking on this is you keep data for its useful business life, right? Or its legally required period, whichever is longer. And since most data isn't implicated by a legally required period, you're down to its useful business life.
That's the problem with the other bucket. Nobody knows what the hell the useful business life is because they don't know what's in there but, but with you know that kind of structured data, being able to say, you know, after 20 months, our customer data is stale. So, we're going to hold it for a rolling 20, and that's what we're going to do our analysis on. And we're going to use the prior and we'll keep the prior analysis, and we'll leverage off of them, but the underlying data we're going to go ahead and let go.
You can make decisions like that far easier with structured and semi structured data, whereas with unstructured that is more difficult. And then I think the other thing about unstructured data. And again, I'm going to like communication data because it tends to be a big source of unstructured data. You know, people are viscerally grabby about their communications right? I need that, I need to be able to go back and look at that. That kind of stuff. Even if they've never done it.
I think it's a harder list too to get the kind of institutional buy in around the unstructured stuff, the structured stuff like there are very few people who are going to fight really hard about whether you keep log file data for two years versus three years versus five years. Unless there's a regulatory requirement, or they've got an actual use case.
And so, I think it's those differences. And you know focusing the value proposition on the structured and semi structured is higher and unstructured data is a morass. I've seen a couple of organizations get super aggressive. They keep email for 30 days. Absent a legal requirement to keep it and that's it.
You can't store it anyplace you can't file it. You can't like, that's it, probably draconian. I don't know if that really works super well, but I understand the impulse because getting people to make that move in the unstructured stuff is difficult and again. You know where's the value proposition in keeping it a lot of times it doesn't tell you a lot. A lot of times it's useful for a lot of things, except for when it was sent or received, right? And the communication that it fostered. That usually then becomes some series of actions that are easier to define downstream. If you need to understand them.
So, for clarification, by structure you kind of include semi structured and somewhat unstructured data that are part of the Jason format databases like you know Mongo DB or Elastic which has become more or even the Amazon Dynamo DB, it's kind of more the common data formats these days than the traditional structured databases that we were used to sort of seeing,
So, I'm guessing kind of putting all of that under the structured bucket. So, you're kind of looking at application data that are sort of formatted in a certain way where you can make sense out of it as opposed to non-application user-generated text chats and communication messages.
Yeah, exactly right. So yeah, I think that that's a good kind of broad way of cleaving the world into two. And that's why I always say, like when I'm having these conversations like I'm referring to structured data kind of writ large, right? Anything where it's existing and sort of that that kind of historic database structure or sort of the more modern structures that are, you know, a little more elastic versus things like communication data, you know application-level data that is really people chattering back and forth to one another or whatever.
Any closing thoughts that you have that you'd like to share?
I do think that increasingly, you know, organizations are seeing real value in analytics, and I think you know tons of studies have been done that show that organizations that have good analytics programs, that are really aimed at the business, do better on average than organizations that don't, right?
And I do think that if you can come in and you can make that argument, I think you're likely to be successful. I do think tying you know this sort of governance conversation to that sort of use case is likely to move the governance conversation forward in a way where having it just as a standalone conversation off by itself is less likely to be successful.
No, I agree. I think you know at the end of the day, we're all, I mean, most organizations are in the business of making money so. Whether it's governance or I mean, while it's essential, even regulatory requirements if you have to meet them, ultimately it has to make a bit of sense. So that makes a lot of sense. Some great great thoughts.
I think that's why I love what you guys are doing right, because, uh, the way you are able to kind of articulate the business value, I think helps a lot in that conversation. So pretty cool.
Thank you, Aaron. Thank you for taking the time.
Thank you for having me.