AI Use Case Canvas
Find where to start. Then prioritize ruthlessly.
The hardest part of AI isn't the tech — it's picking the right first move. Use the grid to brainstorm candidates across every function, then plot them on value vs. effort to see what to do first, what to plan, and what to skip.
Step 1 · Brainstorm
What's possible, function by function.
Each function can go further as you mature — crawl, then walk, then run. Use these to spark your own candidates, then score them below.
Function
Crawl
Walk
Run
Sales
AI drafts follow-ups & call recaps
Lead scoring + CRM enrichment
Personalized outreach agents + CPQ pricing
Marketing
First-draft content & repurposing
Competitor & market research agents
Always-on content engine + attribution
Operations / PM
Meeting-to-action notes
Job-level profitability dashboard
Scheduling & dispatch optimization
Finance
Variance & board-memo drafts
Invoice / AP OCR intake
Rolling forecast + reconciliation agents
Customer Service
Reply drafts (human approves)
Ticket triage + KB generation
Draft-and-approve resolution agents
HR / People
JD & job-post drafts
Resume screening + onboarding plans
Knowledge assistant + workforce planning
Field Operations
Photo / report summarization
Field-data capture + alerts
Dispatch + capacity optimization
IT / Data
Docs & runbook generation
Ticket triage + system docs
Monitoring + integration agents
Exec / Strategy
Research & briefing summaries
Competitor / market research
Real-time ops dashboard + scenarios
Step 2 · Prioritize
Plot them on value × effort.
Add your candidate use cases and score each. They'll sort into four buckets — start top-left.
Do first high value · low effort
Plan high value · high effort
Maybe low value · low effort
Avoid low value · high effort
Add a few candidates to see them sorted. Tip: be honest about effort — data and integration work is usually the real cost.
Adapted from the AI Use Case Canvas in Scaling Up with AI. When you've found your ‘do first,’ a Value Sprint ships it with a guaranteed KPI.