Resource kit
Strategic Intake Canvas
A four-part worksheet for testing the operating foundations around one proposed law-firm AI workflow. Use process information, counts and properly approved examples. Do not place client names, matter details, contract text or privileged material in this worksheet.
Operational guidance, not legal advice. No confidential information is needed. Download the original PDF canvas (original ClickoAI edition, v2, 15 August 2026).
The four-part canvas
Complete the quadrants in order. A blank answer is evidence that the foundation needs work.
- Administrative bottlenecks. Name the trigger, finish line, people, queues, duplicate entry, manual handoffs and non-billable steps. Mark what can be counted today.
- Data readiness. Score 1 to 10. Record where documents live, how they are named, which fields are reliable, who owns the source and how a representative set is retrieved.
- Security and compliance. Map data classes, approved tools, external processors, retention, training use, locations, deletion, human review and the person who can pause the work.
- Value recovery. Use the firm's own volume, lawyer time, turnaround, correction and write-off data. Count supervision and support costs before comparing any option. (The Contract-Workflow Baseline Kit, /resources/contract-workflow-baseline-kit/, collects exactly these five numbers.)
One-sentence working hypothesis: "This workflow may be creating avoidable __________ because __________." Do not turn the hypothesis into a claim. The sections below show what evidence would support or reject it.
Evidence prompts
Use the lowest score you can support. A high score is not approval to deploy.
| Score | Observed condition |
|---|---|
| 1 to 3 | Documents are scattered, naming varies and retrieval depends on individual memory. |
| 4 to 6 | A central system exists, but filing, ownership or metadata are inconsistent. |
| 7 to 8 | The source is controlled and a representative set can be retrieved using agreed fields. |
| 9 to 10 | Ownership, schema, access, quality checks and change control are documented and tested. |
| Question | Evidence or gap |
|---|---|
| Which data classes may enter the proposed route? | |
| Which suppliers and subprocessors touch them? | |
| Where are prompts, uploads, outputs and logs retained? | |
| Can any content train a model or improve a shared service? | |
| Who reviews the output and who can stop the process? | |
| How are data and derived records returned or deleted on exit? |
(The organisation-level checks behind this matrix are in the Enterprise Data Foundation Brief, /resources/enterprise-data-foundation-brief/, and the ten-question systems check at /systems-readiness/.)
Decision record
Turn the canvas into one bounded next step rather than a broad technology programme.
| Field | Working answer |
|---|---|
| Proposed workflow | |
| Named workflow owner | |
| Current monthly volume | |
| Current hands-on lawyer time | |
| Current elapsed turnaround | |
| Current correction or rework evidence | |
| Current write-off or under-recovery evidence | |
| Approved data route | |
| Written human review standard | |
| Decision and reason | Fix the process / buy / run a controlled pilot / stop |
| Reopening or review condition |
A complete canvas does not make the decision. It makes unsupported assumptions visible enough for the responsible people to decide what evidence is still needed.
Next: use the Contract-Workflow Readiness Score at /readiness/ or retain this page with the firm's decision record. If you want the canvas completed with you, request a 20-minute fit call at /fit-call/.
MARGO LEGAL LTD, company number 17322603. Registered in England and Wales. Registered office: 66 Paul Street, London, England, EC2A 4NA. Originally published as a PDF (v2, 15 August 2026); converted to HTML 15 September 2026.