Getting AI To Work by Brennan McDonald

Getting AI To Work by Brennan McDonald

The 10-Minute Diagnostic to Fix Your AI Rollout

Stop tracking performative AI usage and training

Brennan McDonald's avatar
Brennan McDonald
Mar 10, 2026
∙ Paid

Hi there,

Your team completed the AI training. They logged into the platform. Your dashboard shows adoption trending upward. Training completion: 94%. Weekly active users: growing. Reported usage: on track.

Now open the last three deliverables your team produced. The presentations. The analyses. The client work. Can you see where AI changed the output? Not where it was used as an input. Where the output is actually different.

If this is your first article in this series, the framework behind it is here. I have been writing about the friction points that block AI adoption and why leaders consistently start with the wrong one. This week is about the friction point that hides behind your best metrics.

Share

The adoption pattern that looks like success

There is a pattern that appears in most mandated technology changes. The team complies. They attend the training. They log in. They report usage. Everything the organisation measures looks positive. But if you examine the actual work, nothing has changed. The deliverables are produced the same way. The decisions are made with the same data. The workflows are identical to six months ago.

This is compliance theatre. And it is the most dangerous adoption failure because it produces the exact data leaders want to see.

Compliance theatre is not laziness. It is not resistance. It is a rational response to a specific set of conditions. When the organisation measures adoption by activity, the team produces activity. When nobody has addressed what AI integration actually means for the work, the team finds the lowest-effort way to satisfy the metric. They log in. They run a query. They paste something into a prompt. Then they do the work the way they have always done it.

Three conditions produce compliance theatre reliably.

Adoption metrics that track activity rather than integration. Logins, training completions, reported usage. Easy to measure. Easy to perform.

No space for the team to negotiate how AI fits their actual work. When adoption is mandated without input, compliance is the rational response.

Unnamed consequences. Nobody has addressed what AI means for roles, performance expectations, and workflows. Genuine adoption feels risky. Performing adoption feels safe.

In the 5C framework, compliance theatre sits at the overlap of three friction points. Capability: the team lacks access, time, or support to learn how AI fits their actual workflows, so they substitute visible activity. Control: nobody has addressed what AI means for roles, job security, or how decisions will change, so they comply on the surface. Consequences: there is no cost to waiting and no reward for genuinely adopting, so performing adoption is the rational choice.

Compliance theatre is the hardest failure mode to diagnose. It produces the metrics leaders want to see. The dashboard trends upward. The training is complete. The team reports usage. But walk through the actual work products. Open the deliverables. Check the data sources. If the work is being produced the same way it was six months ago, the adoption is cosmetic. The metrics are measuring performance, not change.

The cost you cannot see on a dashboard

The cost is double. First, the organisation invests in tools, training, and infrastructure for AI adoption that produces no return. The work is unchanged. The investment is real. The return is theatrical.

Second, and worse, the organisation concludes that adoption is succeeding. Decisions about scaling, additional investment, and strategy are made on data that reflects performance, not reality. The leadership team reviews a dashboard that says 87% adoption and approves the next phase. The next phase is built on a foundation of compliance. It will produce the same result at greater scale and greater cost.

The organisation is making real decisions based on theatrical data. It does not appear on any dashboard.


If you suspect your adoption metrics are measuring activity rather than integration, that gap is what the AI Change Leadership Intensive is designed to surface. $500, 90 minutes, and you will leave knowing which of your metrics are tracking real change and which are tracking performance. The diagnostic questions below will get you started. The Intensive tells you what to do with the answers. The Intensive includes a full refund guarantee if you do not leave with at least one insight you can act on this week.


Two questions you can answer this week

The diagnostic is simpler than you expect. It does not require a survey. It does not require a new dashboard. It requires looking at the work.

Pick three deliverables your team produced this month. Open them. Can you identify where AI changed the output? Not where AI was used as an input. Not where someone fed data into a tool. Where the output is different because of AI. If the presentations look the same, the analyses follow the same structure, and the client work uses the same sources, the adoption is cosmetic. The team is using AI as an addition to the workflow, not an integration into it.

Ask a team member to walk you through their process for a recent task. At what point did they use AI? What would they have done differently without it? If the answer is “not much,” the AI step is decorative. It exists to satisfy the metric. Remove it and the workflow is unchanged. That is compliance theatre in its purest form: a step that exists for the measurement system, not for the work.

These two questions cannot be answered with words. They require looking at real outputs and walking through real processes. A team that is adopting AI will produce deliverables that look different. A team that is performing compliance will produce deliverables that look the same with an AI step bolted on.

You have looked at the deliverables. You have walked through the workflow. But there is a third question. It is the one that asks whether your own behaviour is producing the compliance you just diagnosed. That question, the Compliance Theatre Diagnostic, and the Accountability Design Template are below, along with every diagnostic tool from the series so far.


You have seen the pattern in your team. The third question below is the one that asks whether you are producing it. The Compliance Theatre Diagnostic takes ten minutes in your next team meeting. The Accountability Design Template gives you four specific replacement metrics your team cannot perform their way through. Paid subscribers also get the full archive: the 5C Diagnostic Decision Tree, Leader Selection Criteria and Redirect Script, Pilot-to-Practice Transition Checklist, Meeting Rhythm Audit with manager scripts for all three meeting types, five playbooks across every friction point, and every tool from every future week. That is a complete diagnostic suite for $19 per month or $199 for the year. Most readers expense it as professional development. The pattern changes when the structure changes.

Keep reading with a 7-day free trial

Subscribe to Getting AI To Work by Brennan McDonald to keep reading this post and get 7 days of free access to the full post archives.

Already a paid subscriber? Sign in
© 2026 Digital Content Operations LLC · Privacy ∙ Terms ∙ Collection notice
Start your SubstackGet the app
Substack is the home for great culture