Client work / Fashion communications agency
The AI was already
on every desk.
How listening to the team changed the agency's plan for AI.
The Original Goal
When we first discussed the engagement, the agency was preparing to introduce Microsoft Copilot. The rollout took place during the assessment. Leadership wanted to understand where AI and automation could make the greatest difference in how the business worked.
I was brought in to assess the agency’s operations, identify promising opportunities, and work out which workflows would be worth building next. The teams were already handling complex, fast-moving work for demanding clients. I wanted to understand where work was getting held up and where AI could genuinely help.
The initial expectation was that, after the assessment and delivery of a customized AI guide, the next phase would involve building workflows.
What the Team
Told Us
Before recommending anything, I wanted to understand how the work moves. 36 people completed an anonymous questionnaire. I then sat down one-on-one with 10 people across departments.
A clear pattern came through. Experienced people spend a large part of their day gathering information before they can apply their judgment. The same facts get organized again and again for different readers.
50%
needed information from several people before they could continue.
47%
manually entered or moved information between systems.
72%
said the task they described took longer than it should.
Interest in AI wasn’t the problem. People were at different starting points, with opportunities that varied by role. Some were experimenting on their own; others needed guidance on using it well and evaluating the output. More than half the respondents named each of two barriers: producing consistently useful results and trusting the output enough to use it.
Why the
Recommendation
Changed
The assessment did reveal workflow opportunities. But the findings also showed that the people who would use those systems needed different kinds of support. Moving straight into building workflows would not address that.
That changed my recommendation. I saw practical AI training as a stronger first investment: giving employees the understanding, confidence, and skills to recognize opportunities within their own roles and improve their everyday work.
I recommended hands-on workshops using tasks from employees’ actual responsibilities, supported by optional drop-in office hours and internal AI champions.
Client relationships, creative concepts, pricing, and final approvals would stay with people throughout.
Explore AI training for fashion brands and agenciesWhat I Built
and Delivered
Built and delivered / Microsoft 365 Copilot
Alongside the assessment, I built the agency its own Microsoft 365 Copilot guide and interactive prompt library, written for how a fashion team works day to day. I delivered the completed interactive resource to the operations director.
Inside the delivered guide / 01
Three Ways to Get Started
A 25-minute Essentials introduction, an extended learning guide with short chapters and step-by-step screenshots, and a prompt library give employees different ways to begin. The extended guide is about an hour of reading.
One of my favorite pieces is a setup step where Copilot interviews each person about their role and recurring responsibilities, then helps write personalized instructions. Those instructions give it a starting context for that person’s work.
Inside the delivered guide / 02
Prompts Built Around
the Agency’s Work
38 prompts across 8 categories cover recurring work, from everyday communication to new business, production, and finance. Staff can search the guide and filter by team or app to find a useful starting point.
Inside the delivered guide / 03
Customize a Prompt for the Task at Hand
Employees enter their task details and see the customized prompt fill in through a live preview. They can then review it and copy it for use in Copilot.
Where Things Stand
The engagement is ongoing. Agency-wide adoption, productivity improvements, and implementation outcomes are not yet established.
Completed and delivered
Assessment and a working resource.
The questionnaire, interviews, assessment, and recommendations are complete. The company-specific Copilot guide and interactive prompt library were delivered to the operations director.
Recommended next phase
Build the team’s capability.
Hands-on workshops tailored to employees’ actual work, with optional drop-in office hours and internal AI champions to support continued learning.
Potential future phase
Revisit what still needs building.
After training, reassess which opportunities still need dedicated workflow design, automation, or connected systems, and which employees may be able to address themselves. This is a possible subsequent phase, not work already implemented.
Your next step
Where could your
team start?
If you’re deciding how to bring AI into your business, let’s talk. You don’t need a plan yet. Tell me what’s taking up your team’s time, and we’ll work out where to start.
Book a complimentary 30-minute consultation Back to Selected Work