Fixing Your OKRs Using AI

Please do not write your OKRs using AI. OKRs should reflect your strategy, your choices, your strong opinion, and your tradeoffs. What you can do is use AI to improve weak OKRs once the strategic intent is yours.

Here are a few ways to do that:

1. Collaborate with AI to create a prompt that reflects your perspective on what makes great OKRs

2. Point AI at an OKR expert you’re following and tell it to create a prompt based on that expert’s advice. (Your OKR Hot Seat with that expert)

3. Point AI at a set of OKR experts and ask it to create a prompt to act as an OKR advisory board, weaving in the expertise/advice from the panel of experts.

4. Read an OKR-related book or take an OKR-related workshop, and take notes in the form of an AI prompt.

Whatever way you create this prompt, I like to save it as context, for example as instructions or a file attached to a Perplexity space. One advantage of this approach is that it works even when using an enterprise LLM, which is often more secure but weaker than public models. Once you create the prompt, you can take it inside the firewall and use it on enterprise OKRs without sharing strategy and priorities with a public LLM.