The framework

The Unified Use Case Discovery Framework

The reason an AI plan fails is rarely the technology. It is that nobody wrote the ideas down the same way, so nothing could be compared and the loudest request won. This framework fixes that: five steps that find the work, eight fields that describe it, and four scores that decide the order.

Part one

The five steps

Two weeks, and under four hours of your time: a 90-minute workshop, three to five 20-minute interviews, and a 60-minute readout. Everything else is ours.

  1. 01

    Establish

    We set up who to talk to and how, so ideas come from the people doing the work, not the boardroom.

    One list of names, one agreed way in. The people closest to a process get asked before anyone is told what to want.

  2. 02

    Discover

    A structured workshop captures every candidate use case across internal operations and customer experience.

    The workshop runs as a capture session, not a wish list. Alongside it we read 90 days of your Microsoft 365 usage — you generate the report; we never read your content.

  3. 03

    Map

    We plot each use case onto your business processes and your customer journey, including the moments after a customer leaves.

    Nothing gets judged in isolation. Every idea is placed where the work actually happens, so gaps between teams and after the sale stop being invisible.

  4. 04

    Structure

    Every use case gets the same eight fields: problem, who’s impacted, sample tasks, touchpoints, business benefit, volume, risks, feasibility.

    One format for every idea, so a request from the warehouse floor and one from the boardroom can be compared on the same page.

  5. 05

    Prioritize

    Each is scored on Impact, Feasibility, Risk, and Strategic Alignment, then placed on the matrix.

    Quick wins first; customer-facing work once the muscle is built. The order is the deliverable, not the list.

Part two

The eight fields

Every candidate use case is written up in the same eight fields — no exceptions for the idea someone mentioned in passing. That is what makes a list rankable instead of merely long.

01

Problem

The specific thing that goes wrong, takes too long, or depends on one person — stated by the people who live with it.

02

Who’s impacted

The roles and teams who feel it day to day, and the customers on the other end when it does.

03

Sample tasks

The actual steps performed today, written out. The work has to be visible, not described in general terms.

04

Touchpoints

Every system and hand-off the work passes through — inbox, SharePoint, forms, queues, the tool nobody admits to using.

05

Business benefit

What changes if the work is done differently: hours recovered, errors avoided, answers given sooner.

06

Volume

How often it happens. A small saving done a thousand times can outrank a large saving done twice a year.

07

Risks

What could go wrong, what data it touches, and which decisions have to stay with a person.

08

Feasibility

Whether the tools and access you already have can do this, and what would have to be true first.

Part three

The four scores

Each use case is scored on four dimensions. Two of them decide where it sits on the matrix; the other two decide whether it moves.

Impact

What it is worth if it is done well — sized, not assumed.

Feasibility

How close your existing tools, data and permissions already are.

Risk

What could go wrong, how visible it would be, and how reversible it is.

Strategic Alignment

Whether it moves toward where the business is actually going.

Part four

Where they land

Impact against feasibility gives four quadrants. You get one page showing every use case placed on it — that page is the argument for what to do first.

Activate Now

High impact, possible with what you already own. This is where the first ninety days go.

Invest & Transform

Worth doing, but it needs groundwork first — data, access, or a process that has to be rebuilt.

Optimize Selectively

Doable today, modest return. We do it when it rides along with something bigger.

Defer

A real problem at the wrong moment. Recorded, scored, and left off the plan on purpose.

Next step

See it applied to your business

45 minutes on the one thing that isn’t working. You leave with a view of where AI applies, whether or not you book the Map.