Recently I was asked to answer some questions for a journalist on how to structure IT organizations. The questions were interesting and I wanted to share my responses here.
What's the best way to restructure an IT organization for maximum productivity?
Of course, the word productivity is an emotive one. Should we strive for productivity or value? Are productivity and value the same? Putting that aside for a moment and focusing on the question is still quite interesting
Productivity is ultimately defined as how much effort is required to deliver the most value. Waste is the ultimate enemy of productivity. The last 2 years of the pandemic have made many of our previous ideas about value and waste outdated. For example, prior to the pandemic, structuring teams around specialist skills, building an amazing planning and resource allocation process (PMO), and putting in place as much automation as possible would for many be the holy grail of productivity. Expensive resources would be optimized. Work would be taken to teams and hand-offs would be tightly managed. The reality is that this approach to productivity relies on certainty. It relies on our ability to know what problems we are solving, what technology we are going to be using, and the skills and availability of teams. The pandemic taught us that the problems, technology, and people are always changing. Of course, in the area of software development, this has been written about for decades and the challenges were first introduced to us in Fred Brooks’ Mythical Man-Month in 1975.
So, what is the best way of organizing in response to uncertainty?
Unfortunately, every situation is different, however, there are some rules that should be considered.
- Keep teams and endeavors as small as possible. For teams, approximately 10 people seem to be the golden rule. For the number of teams, about 10 again seems a clear maximum. But keeping teams as lean as possible reduces the number of hand-offs, communication overhead, and complexity.
- Support those small teams with technically aligned communities (can be called Guilds, Chapters, or Communities of Practice). Because teams are small they will run into technical and business problems that they need help with. By providing a “free” support organization the teams can not only deal with their needs but also develop new skills.
- Provide a clear line of sight to the customer. That customer can be internal or external, but motivated teams are more productive teams. By having a clear relationship with the customer teams will work harder and be more creative in their solutions.
- Have a clear goal and communicate it frequently. In part to support motivation and enable creativity by having clear goals teams become more focused on the outcomes rather than the work. That means that the business must be integrated into the IT organization to provide clearer, outcome-oriented goal setting. And the measures must support that. When incentives are based on work rather than outcomes, people work hard but ultimately may not deliver value.
- Incentivize and empower everyone to be focused on removing waste. This means that space, time, money, and support must be put into the organization to focus on supporting teams removing waste. For example, allow teams a focused Sprint on removing technical debt or increasing automation. Look at all dependencies and hand-offs with a will to remove them. Question all overhead.
- Allow for self-management by providing freedom, guard rails, and goals. It seems obvious, but the best, most productive teams are ones that define how they work and have an environment that supports it. Self-management does not just happen when you say “go and self-manage” it requires careful nurturing and support to build the right level of trust.
In a nutshell, building customer-focused, cross-functional teams that are aligned to business outcomes with a supporting organization drives to more value, more productivity, and a happier set of teams and Stakeholders.
What makes this approach so effective?
I ask a very simple question when people ask me that. If you had a crisis, a big production problem or a massive threat to your business that required immediate action, what would you do? Most people talk about getting the right people, empowering them, making on-the-spot decisions and providing them support. But then they look at their own organization and they see something very different. An organization built around functional areas, systems, and processes with a management model that is focused on reducing risk rather than delivering value.
Of course, they say - that is different. They attest that the crisis model does not scale, is too risky or they do not have enough skilled people.
So perhaps they should focus on fixing that rather than trying to build an organization that is focused inward rather than outward and applies Taylorism in how it treats people. The reason why these organizations work are:
- They accept the challenges of communication and thus construct small teams and try to keep it simple at all times
- They support the notion that empowered people with a clear mission will strive to deliver value
- They understand that the how and why might change so they focus on the outcomes and frequently test
- They are willing to change and do not discourage honesty
- And they care deeply about the happiness of their teams, customers, and stakeholders believing that happiness provides an environment of creativity and partnership
But does this approach really work?
I think that many organizations think of productivity as effort to work rather than effort to value. OKRs, Lean UX, and Design Thinking have challenged our notions of what value is, moving many of us from an output to outcome view of the world. The biggest challenge is not the organization design or even the practices, it is the connection between organization structure and people's sense of belonging and personal value. When I look at organizations I rarely see value streams that are built in response to the needs of customers(products) and a willingness to change. Instead, I see systems that are described in that way but in reality are built around the needs, ambitions, and paranoia of the people within them. I also see organizations that focus so much on the risk that they never really empower teams to feel trusted. How many times have you attended a meeting with 10 people to discuss a $1000 expense? If you build an environment where there is a clear connection with the outcomes and customer, and the teams feel trusted and empowered, then stuff happens. What is funny is that when this does happen, most organizations incrementally stop this success by trying to scale, adding additional governance processes, or moving around staff and incentives. To coin a phrase my Grandmother said frequently about the state of the world. “There is nothing as odd as folk”. :-)