Across our work with more than 75 companies implementing AI, we noticed a pattern: moments of friction where AI could unlock real leverage.

Getting results from implementing AI at your company isn't about chasing the newest tool or feature. It's about identifying friction and using AI to eliminate it.

We created The Friction Framework to fast-track finding high-ROI AI opportunities inside your organization. Take a look at the list below and see where your organization could eliminate friction to create opportunity with AI. Then, meet with your team to brainstorm and discuss together, even 45 minutes spent focusing on this framework will help you find focus.
‍

1. AI Wins

What’s already working with AI that you could build on?

• What is the human friction in the process? How can you flip that so that it becomes leverage?
• Examples: Turn repeat prompts into CustomGPTs or turn manual multi-step flows into agentic workflows.

2. Work Chores

What time-consuming tasks pull you away from high-value work?

• This is the highest-leverage starting point for AI implementation. Work chores are: Repetitive, Cognitive-draining, Necessary but non-strategic. 
• When in doubt: automate the work chores first. 
• Examples: drafting job descriptions, policy Q&A bots, automated onboarding and offboarding.

3. Goals

What would make achieving your goals faster or easier?

• Tie AI directly to your KPIs. If your goal is to increase engagement scores for example, ask: where is friction slowing that down?
• AI works best when attached to metrics, not curiosity.
• Examples: Reports from existing touchpoints (like surveys or performance reviews), AI upskilling hubs.

4. Backlog Projects

What goals remain on the perpetual backlog?

• Every organization has “we’ve always wanted to...” initiatives.
• If something has lived in backlog purgatory for years, revisit it through an AI-native lens.
• Examples: Org scenario modeling, personalized coaching journeys.

5. AI-Native Ideas 

How could existing data or new approaches create value?

• The shift is from “How do we use AI?” to “What becomes possible now that AI exists?”
• Examples: AI note-taking ecosystems, HR + Eng translation copilots.

Once you have your frictions, turn them into AI use-cases with this prompt:

‍You’re an AI-native [your role] at [your org], give me the AI use-cases that would directly solve for these frictions so that I can hit [key KPI] faster or better. Ask me questions about my tech stack, goals, and AI level to optimize your recommendations and help me prioritize. Return the top three as a table with the friction, the use case, who owns it, and what I need to have ready before I start.

Narrow your ideas with the FIRE Diagnostic

The Friction Framework exercise may give you more AI ideas than you can act on. To prioritize, score each one against FIRE:

• Frequency: How often does this happen? Daily beats quarterly.
• Impact: What does solving it free up, in hours or in outcomes?
• AI Readiness: Is the data, access, and process clean enough for AI to work on it today?
• Ease: Could one person build a first version this week?

Score each dimension 1-5, and focus on ideas that are 14+ for quick wins. You can also try out our interactive FIRE Diagnostic on any use case.