Briefing · 14/06/2026

Capacity Destination Mapping: the missing step before team AI automation

Team AI adoption gets easier when staff can see what better work becomes possible. Map the destination for freed capacity before automating the work.

The weakest version of organisational AI adoption is simple: give everyone tools, run a prompt workshop, collect a few time-saving examples, and call it transformation.

That can help individuals. It does not automatically help a team.

The team version has a harder question:

If AI frees capacity, where should that capacity go?

If the answer is vague, adoption starts to wobble. Staff hear risk. Managers hear pressure. Leaders hear a productivity story that may not turn into actual operating improvement.

That is why I think teams need Capacity Destination Mapping before they automate work.

The Adoption Problem Is Not Just Skill

Most AI enablement still treats the employee as the unit of change. Teach the person. Improve the prompt. Share the use case. Increase confidence.

Useful, but incomplete.

Teams are made of roles, handoffs, review points, incentives, trust, customer obligations, internal politics, and workload tradeoffs. A person can save 30 minutes with AI and still create no team-level value if the workflow around them does not change.

Worse, efficiency without a destination can look threatening.

From the employee’s seat, the implied bargain can sound like this:

Help us prove your work can be done faster, then we will decide what that means for your role.

That is not a great adoption message.

It makes caution rational. Some people will use AI privately for safe tasks. Some will avoid it. Some will quietly protect the work that makes their role feel secure. Leaders may call that low AI maturity. Sometimes it is just unclear incentives.

The Better Question

Before asking “what can AI automate?”, ask:

What valuable work is not happening because this team is too stretched?

That is the suppressed-value list.

It might include:

customer follow-up that would help but never happens;
quality checks that are always squeezed out;
documentation that exists only in someone’s head;
onboarding material that keeps being postponed;
process improvement nobody has time to do;
coaching and mentoring that gets replaced by firefighting;
reporting that states activity but never creates learning;
relationship work that matters but is not urgent enough to survive the week.

This is the work AI can make room for.

That changes the adoption story.

Not:

Use AI so we can squeeze more output from the same people.

But:

Use AI so we can reduce low-value manual load and finally do the work we already know matters.

What The 2026 Signal Says

The recent evidence supports this direction.

Microsoft’s 2026 Work Trend Index says AI users are already reporting more time for high-value work and producing work they could not have produced a year earlier. The same report also shows the organisation-level gap: leadership alignment, incentives, and permission to redesign work are still weak in many places.

Glean’s Work AI Index 2026 describes the hidden labour of “botsitting”: the time people spend feeding AI context, checking outputs, rerunning prompts, debugging mistakes, and cleaning up answers. That is the productivity paradox in plain clothes. Individual AI use can create a new invisible job if the organisation does not redesign the work.

The World Economic Forum has been pointing at the human readiness problem too: executive enthusiasm and worker scepticism are colliding. Job-loss fear, distrust, and backstage resistance are not side issues. They are part of the implementation environment.

BCG’s 2026 job-impact analysis makes the practical distinction: task automation does not equal job loss, but a large share of jobs will be reshaped. That means workforce strategy cannot sit downstream of automation decisions. It has to be part of the design.

The pattern is clear enough: AI enablement needs an operating layer between “people are using tools” and “the organisation is better”.

Capacity Destination Mapping is one version of that layer.

How It Works

The mapping does five things.

First, it maps current activity. What work is actually happening? Who does it? How often? Where does it wait, repeat, or break?

Second, it maps friction. Which work is slow, duplicated, error-prone, invisible, manually copied, or dependent on one person?

Third, it identifies suppressed work. What useful work is not happening because there is no time, confidence, tooling, process, or permission?

Fourth, it defines the capacity destination. If AI saves five hours a week, where does that time go first? If it saves twenty, what becomes possible?

Fifth, it writes a role-evolution statement. If AI reduces time spent on one kind of work, what higher-value human work should the role do more of?

That last step matters. Team members need to see a believable path from efficiency to better work, not just efficiency to uncertainty.

A Practical Test

Ask a team lead this:

If AI saved your team 20 percent of its time next quarter, what would improve for customers, staff, quality, or growth?

If the answer is clear, you have a useful direction for automation.

If the answer is vague, the team may not be ready to automate aggressively. They may need to define the destination first.

That is not delay. That is implementation hygiene.

The Line I Would Put On The Wall

Do not automate work until you know what better work the freed capacity will make possible.

That is the difference between AI as private productivity and AI as organisational capability.

Automation Nation Assets

I turned this into a public Automation Nation worksheet and facilitator prompt:

Capacity Destination Mapping worksheet and team-lead prompt

Use it before a workflow automation project, an AI pilot, or a team prompt-training session. The worksheet is deliberately practical: current work, friction, suppressed value, capacity destinations, role evolution, adoption risks, and first 30-day experiments.

The point is simple.

The strongest AI implementations do not just remove work. They reveal the work your organisation has been too stretched to do.

Sources

Was this useful?

Quick signal helps Rob sharpen future briefings.