All insights
Thoughtful AI

Making Room to Lead: AI, Processes, and Better Headcount Decisions

A practical way for busy leaders to simplify work, use AI thoughtfully, and build the capacity their business actually needs.

By Ben Johnson · Clarity2Freedom

Listen to the companion podcast

An AI-generated audio conversation created with NotebookLM.

Your leaders are already carrying more than a full workload. You need the right people in the right seats. There is pressure to hold the line on headcount. And now everyone is being asked to figure out how AI can help.

That is a difficult assignment to give someone whose calendar is already full.

A useful place to start is the work itself. Where does it slow down? Where do people repeat it? Where does a handoff create confusion? What are talented people doing that no longer needs to be done?

Core process implementation gives a leadership team a practical place to answer those questions. AI becomes one of the options you evaluate as you improve the way the business runs.

First, understand the capacity problem

An overwhelmed leader does not automatically mean you have the wrong person. The seat may contain too much work. Responsibilities may be unevenly distributed. Or a capable person may be spending hours compensating for an unclear process.

Those problems require different decisions. Replacing someone will not fix a poorly designed seat. Automating a task will not resolve confusion over who owns the outcome.

In How to Be a Great Boss, Gino Wickman and René Boer write, “Ideally, you should delegate additional activities to get down to 90 percent capacity.” That is a helpful challenge for leaders who have come to accept constant overload as normal.

Capacity includes room to think, lead, develop people, and respond when something unexpected happens. A plan that depends on everyone operating beyond their sustainable capacity deserves another look.

1. Scope the processes and assign leaders

Identify the core processes that make your business work. Agree on where each starts and ends, what outcome it produces, and who will lead the effort to improve it.

That leader is accountable for bringing the right people together and moving the work forward. They do not have to write every word or personally solve every problem.

Start with a process that causes meaningful recurring pain and is manageable enough to improve. A process with repeated rework or difficult handoffs can be a better starting point than the largest process in the company.

Then make room for the work. Decide what can be paused, reduced, or reassigned so this does not become one more responsibility stacked onto an overloaded leader.

2. Create a useful first draft

Use a 20/80 approach: capture the essential steps that produce most of the result. The goal is to give a right person in the right seat what they need to do the work your company’s way.

For each important step, make these things clear:

  • Why are we doing this?
  • What needs to be completed, by when, and by whom?
  • What tools are used?
  • What weekly measurements will tell us whether we are on or off track?
  • Where does the work move between departments, and what must accompany that handoff?

The first draft should be simple enough for people to review together. Detailed instructions can support it where needed, but an exhaustive manual can bury the decisions that matter.

AI can help organize an interview with the people doing the work into a first draft. Use appropriate information in an approved tool, then have those people check the result. A polished description is not proof that the process is accurate.

3. Get the stakeholders together and simplify

This is where the real improvement happens. Bring together the people who perform the process, receive its outputs, and depend on its results. Work through what the right and best way should actually be.

In Process!, Mike Paton and Lisa González put it plainly: “This is where you’ll simplify the process by eliminating, streamlining, or automating steps.”

I would apply that idea through four decisions:

  • Eliminate work that no longer serves a useful purpose.
  • Automate suitable work with AI or existing systems. Some tasks only need a dependable rule or a better connection between tools.
  • Optimize the work people must do by removing unnecessary searching, waiting, duplication, and approvals.
  • Delegate and elevate responsibilities to people who get it, want it, and have the capacity to do it.

Consider a hypothetical weekly report that takes a manager four hours to prepare. Before asking AI to write it, ask which decisions the report supports. Remove sections nobody uses. Let existing systems calculate the numbers. AI might draft a narrative from checked information, while a person remains responsible for accuracy and interpretation.

If preparation drops to one hour but review requires another hour, the net saving is two hours. Include setup and training effort when you evaluate the change. Then decide what those recovered hours will make possible.

This is the connection between process improvement and AI. You clarify the outcome and simplify the work, then choose the technology that fits.

4. Make the process part of how the business runs

A better document only helps if people use it.

Assign ongoing leadership, management, and accountability responsibilities for each core process. Finalize a small set of weekly measurables that reveal whether important parts are on or off track. Depending on the process, those might include turnaround time, errors, rework, or incomplete handoffs.

Train people in a way that helps them understand both what to do and why. Have them practice with real examples and demonstrate that they can complete the work. Sending a document is rarely enough.

Finally, decide how changes will be proposed, approved, documented, and rolled out. This matters especially when AI is involved: people need to know what requires human review and who handles an exception.

Let the evidence inform the headcount decision

The four steps above are my practical application of process and capacity ideas from the EOS® library to today’s AI decisions. They are not a claim that EOS prescribes a particular AI implementation method.

As you improve the work, you may find enough capacity in the existing team. You may discover that responsibilities need to move between seats. Or you may confirm that additional people are needed to deliver the results the business requires.

A few hours saved across several people do not automatically add up to a full seat you can remove. Consider when the work happens, the skills it requires, the demand ahead, and whether the saved time is actually usable.

The useful question is: what combination of people, clear processes, and technology will help us deliver consistently at a sustainable workload?

Start with one process, one accountable leader, and one measurable improvement. Protect the time to do it well. Learn from the result, then expand.

For your next leadership conversation, ask: Where does recurring work consume too much attention? What could we eliminate or improve? And what will we take off someone’s plate so they have room to lead the change?

Books behind these ideas

For a deeper look, purchase and read the original books:

  • Process! How Discipline and Consistency Will Set You and Your Business Free, Mike Paton and Lisa González. Chapters 4 and 5 cover documenting, simplifying, and getting processes followed by all.
  • How to Be a Great Boss, Gino Wickman and René Boer. Chapter 3 addresses capacity and Delegate and Elevate.
  • Traction: Get a Grip on Your Business, Gino Wickman. Chapter 7 connects process simplification with useful technology.
  • People: Dare to Build an Intentional Culture, Mark O’Donnell, Kelly Knight, and CJ DuBe’. Chapter 7 explores how seats and workloads change as a business grows.

For a related perspective on AI, read Mark O’Donnell’s Have You Drawn Your AI Line? and consider where human judgment and relationships matter most in your work.