What we do · Strategy

Discovery Sprint

Six weeks inside one practice area, finding the AI ideas worth building, testing them and ranking the best.

6 weeksOne practice areaA ranked shortlist
The problem

Every practice group has ideas for AI, but they rarely get tested.

The loudest suggestion wins, or nothing happens, and the problems that cost the team most time go untouched.

What we do · Six weeks

Six weeks in one practice

We work with one practice area to map where time is lost, interview fee earners and clients, and test each idea before ranking the ones worth building. We bring our own experience as founders and former lawyers, so you leave with a clear list of options.

  1. 01Week 1

    Learn to think like a product team

    Before anyone suggests a tool, the team learns how product people think: start with the problem, name who has it, and agree what success looks like. Alongside it, an honest briefing on what AI does well today and what is still hype.

    The product lens · Corporate · Example

    "We need an AI tool for due diligence"

    Who has itAssociates running data room reviewThe painTwo days triaging every data room before the real review startsDone looks likeRed flags surfaced before kick-off
  2. 02Week 2

    Map where time is lost

    A workshop maps how work actually moves through the practice: where time leaks, where clients wait and where margin goes. Each complaint is rewritten as a 'How might we' question, which turns it into something a product could solve.

    From the board · Funds · Example
    How might we

    track every MFN obligation across side letters?

    How might we

    check a new LPA against our house terms?

    How might we

    answer a DDQ from past answers?

    Struck off

    A chatbot for everything

  3. 03Weeks 3-4

    Interview fee earners and clients

    Deep-dive interviews with the practice's partners, associates and PSLs, and with the clients they serve. They test our assumptions and show whether people would really use each idea.

    Interviews · IP · Example

    "We check every trade mark watch report line by line."

    Associate · IP

    "Renewal deadlines live in three systems and a spreadsheet."

    PSL · IP

    "I want one view of my whole portfolio, and nobody can give it to me."

    Client · General counsel
  4. 04Week 5

    Test every idea

    Every idea meets the same four questions: does the data exist, would people switch, can it be built inside a quarter, and will someone pay? Ideas that fail are stopped cheaply, on paper.

    Four questions · Employment · Example
    DataSwitchQuarterPaysContract reviewAdvancesHandbook updatesAdvancesSettlement drafting✕-StoppedRight-to-work checks✕---Stopped
  5. 05Week 6

    Rank, plan, present

    The sprint ends with leadership: the ranked shortlist, the reasons behind it, and a 90-day plan for the first pilots, with go and no-go checkpoints.

    The shortlist · Disputes · Example
    Chronology builder8.9
    Disclosure summaries8.4
    Bundle pagination7.6
    90-day plan attachedFirst pilot: chronology builder
The idea board at a Purple workshop: a flipchart headed Idea Board with the Purple mark drawn in marker
From a real Purple workshop
What you get

Three deliverables

  • Discovery report

    The bottlenecks, opportunities and priorities

  • Shortlist of AI ideas

    One page each: the problem, the fix, the saving

  • 90-day plan

    The first pilots, with go and no-go checkpoints

From a real sprint

From debating ideas to building them

An earlier, firm-wide version of the sprint for a fast-growing UK specialist firm. Every idea was scored on value and feasibility, leadership made a clear call on direction, and the 90-day plan had the firm writing code within four weeks of the engagement ending.

We also build what gets chosen, so every idea was scored on what it would really take to build.

Read the case study
Ideas raised
55+
Made the shortlist
6
Weeks to first build
4
After the sprint
Get in touch

Find the ideas worth building

Tell us which practice area is feeling the pressure. We'll tell you whether a sprint is the right first step.