Demystifying “silos,” and why the customer is the cure

May 9, 2018 | Perspectives in Engineering

The word “silo” tends to be one of those shudder-inducing words in any organization. We know we don’t want them — “silos” are a sign that things aren’t going well — but they’re too ambiguous to tackle. So what are they really? And how should you go about knocking them down?

Damon Edwards, Co-founder and Chief Product Officer at Rundeck, regularly speaks at conferences on engineering topics such as DevOps and IT Operations, and has a particular interest in improving how businesses operate. And for him, the ability to break down silos is about focusing on the end-to-end process of delivering value to customers — which, by the way, doesn’t necessarily require you to completely restructure your teams.

“The most critical thing is staying attached to your customers and not having that mindset of ‘We know what’s best, so we’re going to plan this on our own,’” Edwards explains.

A master of knocking down silos, Edwards shared his perspective on how to improve internal operations simply by focusing on the customer experience, and offered his approach to building a learning organization.

Understanding silos — and their symptoms

To prevent silos from forming, it’s first helpful to recognize what they look like in practice. Edwards writes that some people equate silos with teams, but in reality, silos have less to do with organizational structure and more to do with how a group works.

“Part of what causes silos is just human nature,” he says. “We group like with like. Then, there’s the idea of optimization in mass production. You put your developers over there, your operations folks over there, your QA folks over there. In that case, they’re optimizing for their silo.”

In other words, silos don’t necessarily form out of neglect. They really do help certain groups optimize for their functions. But your team’s goal isn’t to be the best firewall rule-changing team west of the Mississippi. The goal is, of course, to deliver value to your customers.

“The customer cares about the end-to-end process. They don’t care about your excellence in your different silos.”

“The customer cares about the end-to-end process,” Edwards says. “The customer cares about what comes out the other side. How quickly does it change? How low is the price? They don’t care about your excellence in your different silos.”

What happens when you cultivate excellence in individual silos is that people work out of context from each other. Team A thinks one way, with one set of terminology, incentives or priorities. Team B thinks a different way, with a different set of incentives and priorities. That’s bound to lead to hand-off and knowledge-transfer problems when, inevitably, teams have to transfer knowledge between themselves to get the work done.

Then, oftentimes the tools and processes won’t align. If your firewall rules-changing team is optimized to be the best in the world, then it’s not optimized to be an effective piece of the whole. Its silo is causing conflict, delay, bottlenecks, and slow feedback throughout the entire organization.

“By optimizing for the silo, you’re not optimizing for the end-to-end flow of value, which is what your business is there to do,” Edwards says. “Historically, people laughed at Toyota. They said, ‘Oh, they’re making their crappy cars for cheap,’ but before too long, they were making better cars for cheaper, and faster. Because, everything they were doing was about optimizing that end-to-end flow of work, not about optimizing silos, which was the traditional way of working.”

The anti-silo: learning organizations

Edwards’ alternative to being a siloed organization is to be a learning one. At Rundeck, he says, they realized that preventing silos wasn’t about putting everyone on the same applications; it’s about what you do around the tools. It’s the organizational dynamics. The processes. The mindsets of how people approach and do their work.

Essentially, are you a learning organization, or are you not?

“I think of a learning organization as ingraining the scientific method,” Edwards says. “Meaning that we have an idea. We’re going to structure a way to figure out if it’s actually true or false. We go and do some work. We evaluate if that disproved or proved what we were thinking. Then we come up with a new hypothesis, and you go forward from there.”

A true learning organization is one that has a continuous improvement mindset — and tight feedback loops — at all levels.

This strategy is also called the plan-do-check-act cycle, or PDCA, in the lean world, and observe-orient-decide-act (OODA) in the military. And it applies on an individual level, too. Your team members can evaluate their day-to-day work through that lens, too. After all, Edwards says, a true learning organization is one that has a continuous improvement mindset — and tight feedback loops — at all levels.

And a big part of that approach means giving contributors the chance to learn for themselves, rather than having to rely on another person or function to do a job for them. “Why not give somebody self-service?” he asks. “Let them do the things they need to do to get their job done, and stay out of their way. Instead of filling out a ticket for somebody else to do the work for them, why not give those people the interfaces to do the work themselves?”

Focus on the customer and amplify feedback loops

Building a learning organization and abolishing silos seems like a daunting task. And the truth is that it may require a significant shift in perspective from your organization and your teams.

Edwards touched on this idea earlier: The best way to become a learning organization is to focus on the end-to-end process of delivering value to customers.

If everyone is focused on delivering value to the customer efficiently and beneficially, then your teams will be less concerned with being the best in their specialized area. Everyone’s there to deliver the product that best serves their customers on their level.

“Look at that end-to-end value,” Edwards says. “Figure out what’s getting in the way. Figure out some countermeasures that you think are going to help alleviate those constraints. Try them. See what happens. Do it again. Do that across organizational boundaries, because we aren’t thinking about vertical silos. We’re thinking about the horizontal value of delivery to a customer.”

He points to Amazon Web Services as a prime example of becoming a learning organization to provide fast value to the customer. “What was so revolutionary was that, as a developer, if you needed something from operations, you would just do it,” he says. “You have an API to get you the things that you need. That means you can move projects along without waiting on others. So those feedback loops get really tight. You can take a lot more turns while your brain is locked into a problem — while it ‘owns’ a problem in its entirety. That’s all important table stakes for these high-velocity learning organizations.”

“We iterate as fast as we can, and at each turn, we look to see what the customer’s actually doing.”

You don’t have to be on Amazon’s level to achieve this effect, though. This is all simply an implementation of lean ideas. You’re failing fast, and correcting fast, to improve your end-to-end process. And by tightening feedback loops, you’re reducing waste (economic and otherwise), reducing risk, reducing variability, and improving quality.

“We have a directional vision for where we want to go,” Edwards says of Rundeck. “Then we look for the next target condition that we want to achieve with our products. We iterate as fast as we can to get there, and at each turn, we look to see what the customer’s actually doing. That’s how we run things.”

In the end, by shifting your focus outward to the customer instead of inward to your performance, you’re getting your product and its improvements to your customers faster and better. That’s what they want. And what they want is what you want.


 

Brook Perry

Brook Perry

Brook is a Marketing Manager at GitPrime. Follow @brookperry_ on Twitter.

Get Engineering Impact: the weekly newsletter for managers of software teams

Keep current with trends in engineering leadership, productivity, culture, and scaling development teams.

Get the Guide on Data-Driven Engineering Leadership

Today, gut feelings in engineering are being replaced with data. By analyzing over 7 million commits from over 85,000 professional engineers, we share how you can incorporate concrete metrics to guide engineering productivity.

Success! Please check your email for your download. You might also be interested in Engineering Impact: the Weekly Newsletter for Managers of Software Teams. Keep current with trends in engineering leadership, productivity, culture and scaling development teams.

Share This