Digital Transformation

Why your company's database could make or break digital transformation efforts

Having data all in one place and easily accessible is crucial to digital transformation success. However, said MarkLogic's Joe Pasqua, many companies struggle with streamlining their database.

Joe Pasqua, executive vice president of products at MarkLogic spoke with TechRepublic's Dan Patterson about how databases can help or hurt digital transformation efforts. Pasqua said that getting company data organized in one place can go a long way towards the success of a digital transformation project. Here's their conversation:

Patterson: Digital transformation is something almost every single company is experiencing right now, and it can range from a strategic conversation to real tactile, technical, implementations ... We're talking about your database today and five reasons your database may be helping or hurting your company's digital transformation efforts. Joe, thanks a lot for your time today. Let's first start a little bit about and talk a little bit about MarkLogic and your expertise with helping company's experience digital transformation.

Pasqua: Great. Thanks, Dan. Right, so MarkLogic is a next-generation database company and we really focus on integrating data from multiple systems. That's intimately tied to the idea of digital transformation because when large companies are trying to transform themselves, usually one of their biggest issues is their data's all over the place. Their most valuable asset, the data about their customers, and suppliers, and vendors, is in 20, 30, 40 different systems.

SEE: Digital Transformation: A CXO's Guide (ZDNet special feature) | Download it as a PDF (TechRepublic)

When they're trying to use that data to make better decisions or implement new operational practices, they're stuck because they end of up having to take this new innovative thing they want to do and tie it to 30 old legacy systems with data in different formats, and rigid structures, and not very nimble. They end up spending, you know, 18 months just trying to figure out how to get to their data and how to organize their data, when they really wanted to do, you know, a quick three-month project to implement some new insight into their customer base.

That's really what we focus on is how do you allow customers, how do you allow companies to easily get at their data, bring it together for operational, or analytical purposes, in a really nimble way? And you know, I just want to touch on one thing you said in your introduction, which is, you know, everybody's looking at digital transformation. The one thing I would say about that is, you know, new companies, younger companies, smaller companies, they really aren't digital natives.

It's the companies that have been around a while that are really looking at, how do we do this transformation and become more agile? They're really, you know, you look at it and the big companies really want to be like little companies. They want to have that agility, and nimbleness, and fast time to results, but they want to do that and at the same time, still take advantage of their scale and all of their assets, their data being their most important asset.

See: 10 bad habits CIOs must break if they want to see digital transformation success (TechRepublic)

Patterson: Joe, I'm glad you articulated that really specific challenge that's shared by a lot of folks who work in enterprise IT and that is, you know, I know my data is all over the place and I know that we want to be nimble, but turning that ship is like the cliché. It just takes a long time to turn the ship around. These are challenges where you can see your data and you can understand the value of merging your data and having a centralized database, but the challenge remains. How do you recommend companies, first, go about cleaning, normalizing, and aggregating data, that like you said, could be all over the place?

Pasqua: Yeah, so you hit on two really important pieces of the problem there. And the first thing I would say is, don't go about cleaning, normalizing, and aggregating your data up front. That's the big mistake a lot of people make and that's dictated by traditional technology. With traditional database technology, you kind of have to do that. You had to say, "Okay, well if I want data from five different systems, or 20 different systems, and every one of those systems organizes its data differently, then I got to stop and create, sort of this, uber schema. This uber data model that's going to reconcile, you know, one model to rule them all that's going to reconcile all my data."

And that means, you end up doing this huge piece of work of normalizing and harmonizing your data before you get any value. And that's just, you know, first of all, it stops innovation dead in its tracks because you've wasted so much time doing that. And the second thing is, companies look at it and it makes them tired just thinking about the problem. It's such a huge thing.

SEE: The top 10 barriers to digital transformation (TechRepublic)

The approach that we advocate is no, don't do that. Figure out what you need to do, and do what you need to do for the result that you need now and start getting a value from that result right away. But, do it in such a way that you can continue to do work over time. You might have GDPR, this EU regulation that's driving you to bring together your data and that's an absolute requirement.

You might be tempted to say, "Well, let me just do, you know, this huge data model effort so I can have all this data together for anything I might need." That is going to take you five years, but the regulation goes live in May. Our model says, no. Use a database and use an approach that allows you to bring your data together, as is, from the original systems. Don't transform it in advance. Harmonize exactly what you need for this problem. Get that problem done. You've still got all your data there. Harmonize what you need for the next problem when you need to solve that problem.

That allows you to do an as-needed approach that lets you get value right away but doesn't stop you the next time you got your next innovation project you want to do. You got this data hub. You got this place that you can go to instead of going back to those 30 different legacy systems. It's really about isolating the pace of innovation of the new stuff from, kind of, the slow pace of the older stuff.

Patterson: Joe, that's great advice and it's results-oriented advice, which I think, you know, speaking only for myself, I know what I see that big task, I just get like, you said, kind of tired. Focusing on results and what I want to accomplish, really helps tackle some of those big challenges, like merging, and cleaning, and analyzing a database.

SEE: 10 companies that are spearheading digital transformation in their industry (TechRepublic)

We teased this a little bit at the beginning and I think we could probably spend all day on each one of these topics, but let's run through those five key flags or those five things you may want to look at, or think about that could help you determine whether your database is a problem or something that's helping you achieve digital transformation.

Pasqua: Great. The very first thing you've got to look at is, is your database making it easier or harder to get data in from all of those different sources? If your database requires you to do all the work upfront, to create these big ETL processes before you get any value, that's a bad sign. That's going to really slow you down.

Step one is, you want to have a database that makes it easy to get data in, regardless of, sort of, the format or style of the data. That's the first thing. The second thing is, okay, now you're bringing all this data together, well, you know, it's your customer data, it's your corporate information, it's, you know, really sensitive, really valuable data, your database better be secure.

You know, there's a lot of next-generation technology that's been developed, that's really cool technology. Not typically aimed at the enterprise and not often times with a focus on, how do I secure this data? How do I make it really really secure, but at the time, you know, you want to be able to use the data? You don't want to make it so secure that nobody can get to it. How do you enable more sharing with less risk?

Hand in hand with that is, governance. You know, you hear about lots of people building data lakes. Take all your data, bring it together, and that's, you know, kind of what we've been talking about. Bring your data together. You know, what we've seen time after time is you bring data together, you put it in this data lake, and you don't keep track of, gee, you know, what agreement did our customers sign when they gave us this data? How am I allowed to use it? What regulations apply?

All of these things that, you know, for big organizations, or for anybody who cares about privacy and governance, you got to keep track of it. A database has to be, not just good at data, it needs to be good at metadata. It's got be able to track all of the things like, where did this data come from? What am I allowed to do with it? Governance is really critical, as well.

SEE: How Sephora is leveraging AR and AI to transform retail and help customers buy cosmetics (cover story PDF) (TechRepublic)

You also want a system that is going to adapt with your organization. One of the main ways that big enterprises are changing these days is where they're running. Today, if you're a big bank, probably most everything you do is on-premise in your data center. But, more and more organizations are looking at, how do I move, at least some of my workloads to the cloud?

If you're a big enterprise, you don't want to move and say, "Gee, I'm going to move to AWS and I'm going to be locked in on AWS." You want to be able to say, "Well, if AWS does a better deal, I'm going to use AWS. If Azure as the better deal, I want to use them. I want to play them off one another." You want to be cloud neutral. You want to have that ability to run wherever it makes sense to run from a corporate perspective.

The other thing that's tied back to the first thing I mentioned is, you want to be able to have a system that can act as this insulating layer. This data hub that really allows you to insulate yourselves from the old legacy technology, because frankly, you know, one of the things people think about with digital transformation is, "Oh, we'll just rip out all the old stuff and replace it." You know, good luck.

You look at how many big organizations are still running mainframes and it's basically all of them. Have they diminished? Sure. But, they're still there. These legacy systems of today that were put in place 30 years ago, they're going to be there 20 years from now. It's a model that says, rip and replace, just as a nonstarter. You have to have a system that integrates well with existing technologies, allows you to migrate off of them, over time, but creates that basis for new innovation that separates you from that older technology. You're right, you can talk about this forever, I just wanted to give you a quick run through.

Also see:

istock-639637280-1.jpg
Image: iStock/littlehenrabi

About Dan Patterson

Dan is a Senior Writer for TechRepublic. He covers cybersecurity and the intersection of technology, politics and government.

Editor's Picks

Free Newsletters, In your Inbox