Breaking down the silos between IT and business [Q&A]
Until the cloud and the idea of big data came along it wasn't unusual for different parts of a firm to have their own information separate from the rest of the business.
But digital transformation initiatives recognize that all information is important to the organization and that silos are increasingly a bad thing.
We spoke to Derek Holt, general manager of Agile and DevOps at digital.ai to find out why there's an issue and what can be done to break down these silos and start to eliminate barriers between business and IT teams.
BN: Silos are something we didn't really hear about until a few years ago, why have they become important now?
DH: The focus has really been on business and technology teams, but there are often silos within IT too, development teams not collaborating effectively with the testing team for instance, or testing not collaborating well with security. These challenges have been there for some time, I think they've only been made more obvious with us all being so globally distributed now.
If we want organizations and business areas to work together we have to make sure that we clearly articulate that is the strategy. Early in my career most big companies looked at their technology investments as an expense, not as a strategic asset. But when you do that you're going to have some misalignment, there needs to be a realization that software is critical to the business. I think organizational alignment is critical and how we align teams matters.
You're seeing more and more folk driving the breakdown of silos by reporting structures and how organizations are formed. I think common language is critical, if one group is speaking about a certain set of key outcomes that they're looking to drive and that doesn't align to the language that's being used elsewhere within the organization then there's a problem.
Lastly the data is a big piece. The more we get to the point where different parts of the organization have their own view of the same data, the more we see those barriers starting to be broken down. If you don't have a strategic alignment, if you don't have organizational alignment, if you don't have terminology alignment, and then if you don't have data alignment, you're going to have challenges.
BN: So this needs a more cross-functional approach between different parts of the business?
DH: I think that's the driver of this whole broader concept. I see companies making this shift where it's no longer the business over here and IT over there, but rather they're coming together around services and products and they have a common alignment. Interestingly enough you see this almost naturally occurred in web companies or for mobile companies, startups over the last 20 years that's just naturally how they've organized themselves. I think you're starting to see that same approach in legacy enterprise environments to drive that same speed and agility and outcomes that they're looking to achieve with technology.
BN: What kind of tools can businesses employ to help break down silos?
DH: A lot of the communication tools that we all use, whether it's messaging or even things as simple as email, those continue to be critical. From my perspective though when it comes to tooling in software development and delivery in business alignment it's really important that you're working as an enterprise and that you're working with tools that can integrate with one another, whether they're from the same vendor or not.
We believe heavily in open integrations even in our company now where we're integrating with hundreds of third party solutions. When you look at all the tools that exist out there the integration of data, of visibility and ideally through the kind of context of the organization those are really key to break down those barriers.
BN: How do you measure whether or not you’re succeeding?
DH: One of the things we talk a lot about is within the software execution layer. We have a set of metrics that have been used for some time, how often are we building? How often are we delivering code? You know that whole velocity of delivery, but what's important to note is that's a great way to measure if we are delivering and developing developing effectively. But it may or may not mean that we're developing the right thing. I believe it continues to be a big emergence area, you see more and more telemetry being able to be pulled off mobile applications and web applications. Things like Net Promoter Score tie back to the actual revenue a transaction has. But this is the really interesting bit, I think next phase is as folks move to value streams. Which is, I make a change I deliver it securely with high quality to the market, and then I get to see whether that innovation makes things better or makes things worse, they rarely stay the same. This is the feedback loop that we're just seeing emerge over the last three four or five years, and I think it's exciting because, as a former developer myself, that was one of the biggest challenges, you delivered and tested the code that made it to the marketplace, but the feedback loop was almost non existent on whether or not that was a good thing for the end customer. I think that's changing in 2021 and obviously will continue to change beyond.
BN: Presumably this will start to extend further down the digital supply chain?
DH: It's interesting, I don't think there's a limit to how how broadly we can continue to mature around aligning to business value. When I say business value what I mean is what your business model is dependent on. Obviously the driver of what you value varies, so if I'm a content provider I care a lot about daily/weekly active users, I care a lot about time in app, I care a lot about click through rate. If I'm an eCommerce business I may care about a different set of things, I might care about percentages, I might care about the number of steps that it takes to get to check out.
I think the key for organizations is first and foremost to define value. Certainly we see some value stream modeling happening more and more, but define value and then once that's defined, however many folks are in the organization or even in the supply chain, makes sense because it allows us to understand our position. Knowing what is valuable and really understanding why we're doing something and what the hypothesis is for the outcome.
BN: How does this tie in with implementing DevOps methodology?
DH: It's a natural progression, if you think of it in five to 10 year increments, if the 2000s were agile the 2010s were DevOps, then I think the 2020s are going to be value streams and that execution layer is a prerequisite to be able to get to the point where you're aligning to the business.