The skills
A plain-language manual of all seven skills - what each one does, what it needs from you, and what it hands back.
A skill is a set of instructions that let this technology carry out one of Thomson Reuters Institute's (TRI's) methods on a department's own answers and documents, then hand back a draft. There are seven of them, one for each cluster of artifacts the toolkit's guides ask a reader to produce by hand. Every skill draws on TRI's published frameworks - the Four Spinning Plates (Effective, Efficient, Protect, Enable) run through all seven, because TRI's whole method is about keeping legal's value balanced across all four rather than defaulting to the one or two that are easiest to count.
Only the first skill below, Design metrics, has actually been built and tested against a sample case - and that one finished skill is the working example being shared with the Institute. The other six are scoped - their inputs, outputs, and the TRI guide behind them are worked out - but the skill itself does not exist yet. Nothing on this page should be read as a finished product.
1. Set ambition Designed
What it's for. Facilitates the department's ambition-setting exercise and a 2030 vision, drawn from TRI's Ambition Setting Explainer and Ambition Setting Exercise guides, together with the Reimagine future legal function exercise. TRI's own exercise is deliberately non-verbal: a leadership team picks 8–10 images or objects that represent the future they want, and the discussion about why those images resonate is where the ambition actually emerges. This is the most conversational and least mechanical of the seven skills - its output is prose a leadership team has to agree on, not a table a script can check.
What you give it. Answers, in conversation, about what the department wants to look like a year or several years out - what a team would put in a retrospective "exhibition" of its own progress, in TRI's framing - plus any existing strategy notes or workshop output the department already has.
What you get back. A drafted ambition statement, a current-capability assessment, a 2030 vision, and a set of priorities - labeled a first draft for the leadership team to agree on, revise, or reject.
- TRI's exercise is built around choosing images, not words - a text-based conversation may not bring out the same thing that visual exercise is designed to bring out. Check whether the draft actually captures shared ambition, or just a leader's individual wish list dressed up as consensus.
- Whether the "priorities" the skill lands on reflect an actual trade-off the leadership team would make, rather than a generic list of good things nobody had to choose between.
- If the five transformation archetypes from TRI's "Reimagine future legal function" exercise are referenced, whether the archetype matched to this department is the right one, or just the most flattering one.
2. Design metrics Built
What it's for. Builds a balanced core metrics framework from TRI's 126-metric Metrics Library, using the method in TRI's Best Practice Metrics Framework for Legal Departments and Introduction to Metrics Library guides. This is the only skill built so far, and it is the heaviest of the seven - it reads the full Library, checks candidate metrics against what the department can actually collect, and balances the selection across all four Spinning Plates rather than letting it default to cost and speed.
What you give it. The department's business objectives (the company's goals, not legal's restatement of them), whatever metrics it reports today - clean, messy, or none at all - a candid account of what its systems can reliably produce, who the reporting is for, and targets if any exist.
What you get back. A drafted core metrics table of 8–12 metrics, roughly 2–3 per plate and covering all four plates, each one drawn only from TRI's Library and carrying a plate, the business objective it serves, a data source, an owner, and a target. Alongside it: a measurement definition for each metric, a gap list of what the department should measure but currently can't, a plate-coverage summary, a list of open items to resolve, and a five-category maturity self-assessment against TRI's best-practice table.
- Whether each metric was mapped to the correct Spinning Plate. The skill is meant to take the plate straight from the Library's own column rather than guessing from the metric's name - check that it did, especially for metrics that sound like cost or efficiency but that the Library actually places elsewhere.
- Whether the 8–12 metric, 2–3-per-plate rule was applied as TRI intends - as a general framework guideline to balance the whole picture, not as a rigid quota that forces in a weak metric just to fill a plate.
- Whether a metric TRI would consider essential for this kind of department is missing entirely, particularly on the Protect plate, which rarely ladders up to a stated business objective and is the plate most likely to come back empty or thin.
3. Design AI metrics Designed
What it's for. Measures AI adoption across all four plates instead of time saved alone, using TRI's AI Success Metrics Worksheet and its Best Practice Guide. Kept as a separate skill from Design metrics because it has its own worksheet, its own narrower metric count, and its own reinvestment logic: TRI's argument is that measuring AI purely on efficiency invites the conclusion that if AI removes a quarter of the effort, a quarter of the budget can go too - so the worksheet pairs metrics with a plan for where the freed-up time and money actually go.
What you give it. The department's AI initiatives, the benefits expected from each one across the four plates, which stakeholders care about which benefits, and which of TRI's reinvestment options (skills development, higher-value work, relationship building, proactive risk management, innovation) fit the department.
What you get back. A drafted AI metrics table of 3–5 core metrics - this is a smaller, separate count from Design metrics' 8–12, on purpose - each one covering at least three of the four plates, tied to a named stakeholder, and feasible to collect. Alongside it: a reinvestment plan with named owners, and a draft AI success narrative.
- Whether the 3–5 metric rule and its "at least three plates" criterion were kept distinct from Design metrics' 8–12 / 2–3-per-plate rule - these two rules belong to two different TRI worksheets and are easy to blur together.
- Whether the reinvestment plan actually reflects TRI's five named reinvestment options and names an actual owner, rather than reading as a generic "we'll be more strategic" statement.
- Whether the selected metrics genuinely "tell a story beyond efficiency alone," as TRI's own selection criteria require, or whether the draft has quietly defaulted back to time-saved as the only measure.
4. Design stakeholder research Designed
What it's for. Scopes a Voice of Stakeholder research project before any data is collected, using TRI's Scoping & Planning Guide and Question Guide. It is deliberately separate from the skill that analyzes the results, because the inputs are completely different: here, there is a stakeholder list and a set of open questions; the results don't exist yet.
What you give it. What the department wants to learn, who its stakeholders are (executives, board members, business units, high-volume partners), and any constraints on method (survey, interview, or a mix).
What you get back. A drafted set of research objectives and a stakeholder map, a recommended method, a questionnaire or interview guide built from TRI's question bank, and draft invitation and follow-up communications.
- Whether the questionnaire respects TRI's stated limits - a maximum of 20 questions for a web survey, 10–15 for an interview - rather than growing past what respondents will actually complete.
- Whether the research objectives were narrowed to TRI's recommended 3–4 maximum, rather than left as a long list that tries to answer everything and answers nothing well.
- Whether the stakeholder map actually reaches the people whose concerns matter most to this business - a specialist or regulatory voice, for instance - rather than only the stakeholders who are easiest to reach.
5. Analyze stakeholder feedback Designed
What it's for. Turns collected survey or interview data into findings and a plan, using TRI's Data Insights Guide and Action Planning Workshop guide. Separate from the research-design skill above because now there is actual data to work with, rather than a plan for collecting it.
What you give it. The raw responses - scores, ratings, and open-text comments - and the original research objectives from the scoping stage, so findings can be checked against what the department actually set out to learn.
What you get back. A drafted findings synthesis (a performance dashboard, a look at the spread of responses rather than just the average, gaps between expectation and experience, and the recurring themes in open comments), a prioritized action plan split into quick wins and longer strategic initiatives, and draft communications back to stakeholders.
- Whether an overall average score is quietly hiding a sharp split between two groups - TRI's own guide is explicit that looking only at the mean is a failure mode, so check the draft actually shows the spread, rather than only the headline number.
- Whether a difference between groups that the draft calls meaningful actually is - TRI's rule of thumb is roughly twice the standard error before calling a gap genuine - or whether it's reading noise as a finding.
- Whether the themes pulled from open-text comments are backed by actual quotes from the data, not smoothed into something that sounds plausible but wasn't actually said.
6. Build the value story Designed
What it's for. Distills a finished metrics framework into a board-ready slide and narrative, using TRI's Tell Your Value Story: One Slide Story guide and Writing for the Board. It is downstream of Design metrics (and Design AI metrics) - it consumes a framework that already exists rather than picking metrics itself, because provenance is especially important here: a slide travels further than any other artifact this toolkit produces.
What you give it. A completed metrics framework (from the Design metrics skill, or supplied directly), and a sense of the audience - board, CEO, or an internal variant.
What you get back. A drafted one-slide board story covering all four plates with 2–3 hero metrics per plate, an internal legal-department variant, and a written board-ready narrative.
- Whether every figure on the slide actually traces back to a metric in the framework that has a named source - a number with no traceable origin should never reach a board deck.
- Whether a metric that was flagged as uncertain upstream (for example, one with no trustworthy year-over-year baseline) still carries a visible caveat on the slide itself, rather than being smoothed into a clean-looking number.
- Whether the 2–3 "hero" metrics chosen per plate are genuinely the most compelling ones for this particular board, rather than whichever metrics happened to be first or easiest to pull.
7. Audit and advance Designed
What it's for. Two jobs. First, TRI's published audit: assess a department's current metrics against best practice and prioritize the fixes, using the Audit Your Approach best-practice guide and its Gap Analysis Worksheet (five categories - metrics coverage, data collection, engagement, reporting, review - each scored 1–5). Second, a reconstruction: TRI's own toolkit links a seventh module, "Maximize your impact," meant to cover scaling success across all four plates once a department has been through the cycle once - but that page on TRI's site is currently empty. This skill's second half is my best inferred reading of what that section is meant to do, built from TRI's own audit and reinvestment material, offered as a question rather than a claim.
What you give it. An existing metrics framework (ideally the output of Design metrics), or a description of what the department currently measures and how it operates day to day.
What you get back. A drafted gap analysis scored against TRI's five categories, with the current state and the specific gap named for each; and a prioritized improvement roadmap, including a draft attempt at the "scale what's working" logic that would fill TRI's missing seventh module.
- Whether the 1–5 score given in each of the five categories actually matches TRI's rubric and the descriptions in the worksheet, rather than a plausible-sounding guess.
- The "Maximize your impact" reconstruction is the least grounded part of any skill on this page, since TRI never published the content behind it - check whether my reading of what that module should contain is actually close to what TRI intended.
- Whether the priority levels attached to each gap reflect actual business urgency for this department, rather than simply following the numeric order of the scores.