Lisa Guan's Innovation World

As the author of 20 Principles of Practicing OKR, I’ve written—and helped others write—countless OKRs. When ChatGPT and other AI tools became popular, I started experimenting with them in my own OKR process.

To my surprise, AI turned out to be incredibly useful. Not because it writes the “perfect OKR” (spoiler: it doesn’t), but because it speeds up the messy parts, makes ideas sharper, and helps teams see what they might otherwise miss.

Here are some of the ways I’ve found AI to be a real game-changer in OKR writing.

1. Save Time on Drafting and Iteration

Teams often waste hours (or even multiple meetings) wordsmithing OKRs. AI can create a decent first draft in seconds, which frees up time for the team to focus on what really matters: discussion and alignment.

Example: Instead of holding three meetings to get the wording “just right,” you spend one session refining AI-generated suggestions.

2. Clarify and Refine Wording

We’ve all seen OKRs with vague objectives like “Improve customer experience.” The problem of this kind of OKR is that everyone will interpret that differently.

AI can take fuzzy goals and reframe them into something sharper and measurable.

Example:
Vague: Improve customer experience
AI-suggested: Increase Net Promoter Score from 50 to 65 by Q2

3. Shift from Tasks to Outcomes

A classic mistake: writing tasks as OKRs. For example, “Launch feature X.” That’s an output, not an outcome.

AI can help reframe these into results that actually show impact.

Example:
Task: Launch feature X
AI-suggested outcome: Reach 40% adoption of feature X within 3 months

4. Align with Company Strategy

AI isn’t just good at writing words—it can also analyze context. By pulling from company strategy documents, past OKRs, and team goals, AI can suggest OKRs that clearly ladder up to the bigger picture.

Example:
If the company-wide OKR is “Expand global market share,” AI might recommend:

• Marketing: Launch localized campaigns in 3 new markets
• Product: Adapt features for regional compliance in Asia and Europe

5. Bring in Benchmarks and Best Practices

Sometimes teams write OKRs in a vacuum. AI can surface examples from OKR playbooks, case studies, or even other industries to broaden your perspective.

Example:
Your fintech startup writes: Grow user base by 20%.
AI suggests adding:

• Pass PCI-DSS compliance audit
• Achieve <1% fraud rate on new accounts

These additions make the OKR more comprehensive, not just growth-focused.

6. Encourage Stretch but Realistic Targets

Setting ambitious goals is part of the OKR philosophy. But how ambitious is “too ambitious”?

AI can analyze historical data and suggest stretch targets that are challenging but achievable.

Example:
Last quarter you improved NPS by +5. AI might suggest aiming for +10—not an impossible +30.

7. Add Missing Key Results for a Complete Picture

Humans often stop at the obvious output. AI reminds you to think about success more holistically.

Example:
Team writes: Release AI chatbot.
AI adds:

• Maintain customer satisfaction above 90% post-launch
• Reduce support ticket volume by 20% within 2 months

8. Balance Across Multiple Dimensions

Good OKRs aren’t one-dimensional. They should balance business, customer, process, and team perspectives.

Example for a product launch:

• Business: $2M revenue from the new product
• Customer: 4.5+ app store rating
• Internal: 95% uptime in first 3 months
• Team: <15% overtime reported by engineering

AI also helps avoid “positive-only” OKRs by adding guardrails.

Example:
Your team writes: Grow sales 30%.
AI adds: Maintain churn below 5% and Keep gross margin above 60%.

Some final thoughts

AI won’t write perfect OKRs for you. But it does act like a second brain—sharpening your ideas, surfacing hidden success factors, and saving you time in the process.

Think of AI as giving you a “good enough” starting point. From there, your team can refine, align, and build OKRs that are not just faster to draft, but also more balanced and comprehensive.

In other words, less time wordsmithing, more time driving real outcomes.