Test Data Management Takes Center Stage in the Race to Innovate: An Interview with Chris Fuller

[interview]

by Delphix

Summary:

In this interview, Delphix vice president of worldwide customer success, Chris Fuller, explains why test data management has become so relevant in software today. He defines the term, explains its business impact, and details how your company can use Delphix for test data management.

Josiah Renaudin: Welcome back to another TechWell interview. Today, I am joined by Chris Fuller, the vice president of worldwide customer success at Delphix. He'll be speaking on test data management. Chris, thank you very much for joining us today.

Chris Fuller: Hi, Josiah. How are you?

Josiah Renaudin: I'm doing fantastic. And to kick things off, before we really dig into test data management itself, can you give us a brief overview of your experience in the industry?

Chris Fuller: Sure, so I've been in IT for a little over twenty years now. Started my career back with Anderson Consulting and Accenture. Like most of the folks with those organizations, a lot of coding, a lot of testing, so it's a space that I've grown up in. I spent a little over seven years with Oracle, running their North American CRM practice, mostly focus on strategic accounts. Again, focused on big technology implementations, complex integration programs. For the past three years, I've been with Delphix, where I came on board to establish and grow the customer success function globally.

Josiah Renaudin: I think it's always important that, if we're going to talk about something like test data management, we should probably explain exactly what it is in this context. So, the big question: What is test data management?

Chris Fuller: Yeah, so test data management is quite simply the function of creating the test data that testing teams need in order to execute their test plans appropriately. What we find in most organizations is that this is a centralized capability that really serves as a data hub, if you will, to provide data to all the different teams that need it.

Josiah Renaudin: ... and, beyond just defining it, I think a lot of times we define things, and we're like, "This is why it's important, because it does this," but we need to know more about how it's impacting companies, how it's impacting software, how it's impacting business, so ...

Chris Fuller: Yeah.

Josiah Renaudin: ... why, in your mind, is test data management becoming such a growing area of focus?

Chris Fuller: Yeah, I think it has more to do with the macro trends around customers and companies really racing to innovate. I think this is true across every geographic territory, it's true across every major industry, but every company is out there, and they're facing new competitive pressures that they really haven't faced before. In our digital economy, that really means in many cases getting applications to market faster, and getting new application capabilities to market even faster.

You simply can't do that without testing. You have to test those capabilities before they're released, and not getting good quality test data quickly can be a real impediment. What we're finding is a real focus on test data management, an evolving view of what test data management actually means, and all that's really around helping companies unlock the innovation that they have and unlocking that innovation in their field.

Josiah Renaudin: I feel like there's always, even when I came into this job about two, two and a half years ago ... There's always talk about traditional versus the new-schoolers. How people have done it before, versus how people are doing it now—in software, in technology. It's always changing, so you have to change with the times. How have organizations traditionally approached test data management, and how is that changing as the industry itself is changing?

Chris Fuller: Sure. I'd say test data management traditionally has really focused on the aspect of synthetic data creation: I've got a new application that I need to bring to market, or I've got new capabilities, and so I need somebody to go create that data for me. That's been the primary focus, historically. What that has caused, though, is that different project teams, application teams, have had to deal with all of the other data needs themselves. If they need fresh copies of production, if they need to mask their data, if they need to create different types of subsets, that responsibility has really fallen to the individual project teams themselves.

I think the pressures of things like data security across enterprises, the pressures to continually do things more efficiently, more effectively, are really changing the notion of test data management to where these functions need to provide a much wider array of services. It needs to include things like synthetic data creation, but it also needs to address all these other key areas as well.

Josiah Renaudin: So often we see something have success—again, I always go back to agile—where it's like, "Okay, agile's working. Let's go do that. I see another team doing it. Let's incorporate it in our team, and everything is going to be great ... "

Chris Fuller: Yeah.

Josiah Renaudin: ... but I feel like people tend to, maybe not ignore, but they kind of overlook the issues that might come about, the issues in putting it in, and not just actually instituting it, but maintaining agile and things like that.

Chris Fuller: Yeah.

Josiah Renaudin: So, what challenges do organizations face as they try to build out robust test data management capabilities in response to growing application demand?

Chris Fuller: Sure. I think the key issue is really fully recognizing the very broad set of groups that need to be served. As I would say previously, things were somewhat myopic and focusing on specific testing needs, but now there's really a wide group of testing that needs to be addressed. You've got things like unit tests that are happening in development. You've got different flavors of system integration testing. You've got performance and stress testing. You've got user acceptance testing. You've got automated regression testing.

All of these different types of testing really need to be taken into account by modern test data management teams, and further, you also have to think about all the different actors that are also involved. You have employees, you've got contractors, you've got consultants, you've got parties working onshore, you've got parties working offshore. It's really a very diverse set of situations that test data management teams need to address, and they need to think about that very comprehensively so that they're not just simply a one-trick pony that's only solving a very small piece of the overall testing problem.

Josiah Renaudin: You mentioned all these different moving parts, and that's more and more having to do with modern business, where people are remote, people are in different countries entirely. There's so many different moving parts. Especially on a software team, there's the testerrs, the developers, all these different people.

Chris Fuller: Yeah.

Josiah Renaudin: How do you get everyone on the same page? I know that's not an easy question to answer, because everyone's different; but do you kind of look at a certain script where you're like, "All right, this is how we get everyone on the same page so that test data management can function properly"?

Chris Fuller: Yeah, see, you really don't have to get them on the same page per se, but you've got to recognize that there are those diverse needs that are out there, and they typically do fall into different types of patterns. What we typically see, teams need access to production data. They need access to master production data. They need access to synthetic data subsets and, increasingly, they need to group these things together as well; so if you're doing things like very complicated integration testing, you need a wide set of data that you can treat as a single group, as a single package.

In some cases, you're even aggregating these different types of data sets together. I think if you kind of reduce it down and look at the base of it, there are some very core patterns that we see, and it's through the efficient management and distribution of those types of data that test data management teams can be maximally effective.

Josiah Renaudin: And let's continue to look at the present and what's coming, because again, I know it's important to understand what's working now, but what are we going to have to understand and institute in the future. What are some of the test data management technologies that businesses are leveraging today? To branch off of that, how are they changing, and what does the future of this technology look like?

Chris Fuller: Yeah, so I'd say the focus previously has really just been on data creation, so there are a lot of tools that do that sort of thing. There are even new market entrants that have some pretty interesting capabilities, but the shift, in order to meet this diverse set of needs that are out there, the technologies really need to keep pace with that. Now the focus isn't so much now just simply on creation, but now it's also focusing on security, focusing on efficient distribution of that data across an entire enterprise. It's also a focus on enabling self-service.

With the rise of DevOps, with the rise of agility, with the focus on automation, it's important that individual constituents, individual developers and testers, that they start to be able to self-service their own needs as well. There's a really big shift in the technology underpinnings of these capabilities, to where we don't simply want a centralized team that's there in a fulfill, in a request fulfill sort of model.

We really want a robust set of data capabilities that are there, available for developers and testers to really self-service themselves in a way that's highly orchestrated, a way that's highly automated, and has all of the right security controls across the enterprise to make sure the data is protected.

Josiah Renaudin: For these organizations that are looking at really building out new test data management processing tools, what criteria should they use to evaluate their approaches? It can be difficult to, there may ... You may be able to align all of these new approaches and say, "Okay, which is the right choice for my team, my organization?" What criteria are you looking for?

Chris Fuller: Yeah. I'd say one of the key elements that's really emerging is the notion of data virtualization. If you're going to be handling this amount of data, if you're going to be distributing this amount of data across an enterprise in all of these different ways, you really can't do that efficiently and effectively, and in a cost-effective manner, without significant virtualization. That's one key capability that test data management teams really need to look at closely and understand.

They also really have to look at integrated masking capability, so if data is to be distributed across an enterprise, if we are to service this wide group of constituents that include employees and contractors and third parties, onshore and offshore, data security is a really important element of that. Then having self-service capabilities also built in is really important.

I think, as teams are out evaluating different technologies that they want to bring in, finding the sorts of integrated platforms that are built with all of these capabilities in mind really help accelerate getting the sorts of enhanced capabilities in place, faster, rather than having to build them out in a very piecemealed fashion that requires a lot of administrative and overhead overtime.

Josiah Renaudin: Let's continue to hammer away at that "Why?" What does this do for the team, what does this do for individual people? What kind of impact does high-performing test data management have on the QA teams, and the big question, again, why should test data management matter to consumers of test data?

Chris Fuller: Yeah. If you put yourself in the shoes of an average developer or a tester, getting the data that they need can be a real impediment to them producing their deliverables at a very fast pace and producing them with very high quality. Having unfettered access to the specific types of data that you need, when you need it, is really a key to unlocking that productivity and unlocking that innovation potential. You don't want a testing team that has to sit around and wait hours, days, or even weeks to get the data that they need. They need that on tap. They need it immediately available, and that means that they can get to work on the things that are truly value-add, versus just spending a lot of that time in wait mode, or trying to find other productive things to fill their time.

Josiah Renaudin: Now that we've defined it… we've talked about why it's important, where it's going, we got to, of course, talk about Delphix, the company you work for. Why is Delphix transformational for test data management?

Chris Fuller: Yeah, so Delphix was purpose-built to solve many of these data problems. Delphix is a data agility platform, and it focuses on being able to consume different types of data, whether that be synthetic data sources, production data sources, or subset data sources, and then basically making that data available on tap. By virtualizing the data that we bring in, we can drive massive benefits across both infrastructure and storage; but the real power is in the ability to get that data to developers, to testers, in a way that they can actually start to self-service their own requests. What's really fascinating is, once you start to virtualize data, you can do some things with it that you can't do with traditional physical data. Now a tester can run a test, they don't like the results, they can refresh that data, they can rewind it to individual points in time, they can bookmark it, they can branch it.

They can start to do things and pivot in ways that were just previously not even possible. Delphix provides all of these sorts of virtualization and agility types of benefits to the developers and testers that need it. It really unlocks a tremendous amount of productivity from the groups that are able to benefit from it.

Josiah Renaudin: Let's give the listeners and readers an example of a success story that Delphix had. Can you talk through an example of a business using Delphix for test data management? What were they trying to change, and most importantly, what were the results?

Chris Fuller: Sure, absolutely. Delphix is working with a significant number of the world's leading enterprises that are really focused on these sorts of topics. I'll highlight very briefly a leading healthcare provider that we have here in the U.S. They, like many other organizations, were fundamentally transforming how they looked at development, how they looked at testing, how they looked at test data. The first step that they took was to basically virtualize all of their development and their testing environments. This drove massive amounts of infrastructure and storage efficiencies, but it also created a tremendous amount of power for the individual developers and testers. They also provided these developers and testers with all the sorts of self-service capabilities that I talked about so, in the hands of those developers and testers, the ability to create new environments, to refresh environments, to reset them, to bookmark them, to share the data ...

The net result is that that team, that organization, has gotten significantly faster in getting brand new products to market. That's driven their competitive position overall, and it's helped them to really grow as a company. In fact, they've more than doubled the size of their already substantial business by unlocking this type of innovation. That's really the power of Delphix and TDM, and that's the power of really unlocking innovation at speed.

Josiah Renaudin: That's a fantastic example. If there's a team, organization, or anyone who's listening right now, who would want to get in contact with Delphix to institute and learn more about test data management, what's the best way to do so?

Chris Fuller: Best way is to go to our website, www.delphix.com. You can see all kinds of great information, great use cases, great stories about the things that we're able to unlock. There's easy ways to get in contact with us to get more information.

Josiah Renaudin: Well, thank you so much, Chris. It's very obvious you know this stuff front and back. Anyone who's interested in test data management should visit delphix.com, and I am looking forward to speaking with you again about a different topic in the future.

Click here to take a short survey to let us know how your organization manages test data.

Chris FullerChris Fuller joined Delphix in 2013 to establish and operate its worldwide Customer Success organization. Chris and his team of experts partner closely with Delphix’s most strategic customers to facilitate the transformation of their application development, testing, and operational processes with agile data. He has over 20 years of IT experience and previously served in executive and management roles at both Oracle and Accenture. Chris has personally worked with the world’s leading financial, communications, energy, healthcare, and technology companies to drive innovations that accelerate time-to-capability. Before starting his career in professional services, Chris earned his bachelor of science in Industrial and Manufacturing Engineering from the Florida State University.

About the author

Upcoming Events

Oct 13
Apr 27