We have too many connections in the way we work now.
By a connection, I mean any time one person needs to connect with another person to get work done.
Leaders create these connections all the time, often without realizing it.
- Leaders change how work is done (even a little), requiring people to connect with someone new.
- Leaders push employees to be more responsive.
- We become so specialized that “someone else” is always responsible for the next step.
A single connection has a marginal cost. Needing to talk to someone or hand something off is no big deal—you have time. But every one of these connections costs time and communication.
You have to make time to communicate using email, phone, chat, meetings, or in-person conversations. If it’s a phone call, meeting, or in-person discussion, you have to find a time that works for everyone. Then you have to have the conversation, clarify anything that wasn’t understood, and sometimes restate it in another form. If it’s chat or email, there is often miscommunication, more clarification, or a long back-and-forth before work can continue.
The problem is when these connections reach a tipping point and start costing more than they are worth.
I keep thinking about work that was done in the past, before computers. Somehow we often got it done with fewer people and, in many cases, faster. I’m not convinced computers made the work itself slower, but they made creating new connections almost effortless. It’s easy to add another approval, another meeting, another chat, another owner, or another handoff. Each one seems inexpensive on its own, but they accumulate over time.
I think a key problem is that we have become individually over-specialized. We can only do a limited number of steps before we have to hand something off to someone else.
In some organizations, it becomes even more complicated because every project follows a slightly different process. Each activity may involve interacting with a different group of people, requiring a different set of handoffs. The work itself becomes less complicated than navigating the network of people needed to complete it.
We have over-indexed on specialization.
This reminds me of Brooks’ Law from The Mythical Man-Month. Fred Brooks observed that adding people to a late software project often makes it later because every new person creates more communication and coordination. Existing team members have to answer questions, provide context, and keep everyone synchronized. The additional capacity comes with additional communication overhead. I think specialization can create a similar problem. Every new owner or department may improve one task, but it also creates another connection that has to be managed.
We are trying to be an assembly line without the physical limitations.
On an assembly line there are only two goals: keep the line moving and complete each step. In that environment, specialization makes sense. One person installs the door, another installs the window, and so on. There is no handoff process for each employee. They don’t need to talk about every car—they simply perform their task.
If a leader sees an inefficiency, they can change the process and verify whether it helped by asking a simple question: Did the assembly line go faster?
Notice, too, how changes are made. There is usually a formal process involving leadership, industrial engineering, the assemblers, and others. The change is implemented, verified, documented, and standardized before it becomes part of the process.
Knowledge work is different.
One of the process improvements leaders often implement is specialization. They make an individual or department the sole owner of a task. Now there is a handoff into that group and another handoff out of that group. There is more waiting, more communication, more emails, more chats, and more meetings.
If that work were spread among everyone, much of that coordination would disappear. A person does not need to email themselves. They don’t need to have meetings with themselves. They don’t need to explain things to themselves.
Specialization has benefits, but those benefits need to be weighed against the administrative overhead that comes with it.
Managers seem to pay close attention to the task itself and very little attention to the administration that surrounds it.
One advantage of knowledge work is that we can change our processes much faster than an assembly line can. But we also have far more people making those changes, and almost none of them are verified to determine whether they actually help the organization produce more.
We don’t have an assembly line where we can simply watch it move faster or slower.
Leaders can layer on process changes, more connections, more specialization, and more activities with no verification that the organization is actually producing more. Over the years, you see processes cycle back and forth as each new leader changes them again.
Maybe we need to make process changes harder in large organizations.
Before adding another approval, another owner, or another handoff, we should answer one question:
Does this help the organization produce more, faster?
The answer has to be measured at the organizational level, not at one department.
If one department moves faster but creates a pile-up at the next step, the organization as a whole is actually slower.
Eventually we end up with so many of these connections that much of our work becomes servicing the connections rather than doing the work.
Then managers start talking to employees about how quickly they need to reply to emails or Teams messages.
Sometimes I wonder if this is simply the nature of large organizations.
Maybe we need a way to measure a person’s work.
How much of their day is spent producing something?
How much is spent talking about producing it?
If communication becomes the majority of the work, maybe that’s a signal that something is wrong.
I think many leaders unintentionally count communication as productive work because they don’t distinguish between doing the work and coordinating the work. That drives the wrong behavior.
I also get the sense that the problem is worse than senior leadership realizes.
As leaders climb the organization, there is a natural filtering of these communications. They aren’t interrupted by every Teams message, every approval request, or every small coordination issue. Yet they continue putting processes in place that create those interruptions for everyone else.
Cal Newport makes a similar observation in Slack Is the Right Tool for the Wrong Way to Work. Slack isn’t necessarily the problem. It is a tool that works well for organizations built around constant coordination.
I don’t know about you, but all of this jumping between conversations makes me miserable.
When I get a little time between meetings, I immediately try to get started on a larger piece of work. Just as I’m beginning to get into it, someone sends me a Teams message. It pulls me away, and by the time I return I’ve lost my train of thought.
All day I think about the one project that needs a large block of uninterrupted time.
By the end of the day I’m exhausted, but I can’t point to one meaningful thing I accomplished.
Trying to juggle all of these conversations while also trying to do deep work is miserable.
The work that actually requires sustained thinking gets pushed into evenings and weekends because the workday has become consumed by coordinating the work.