How to Use AI to Set Better Team Goals with OKRs

I asked an LLM to help me set business goals for the year. It gave me a 47-item to-do list that made me want to close my laptop and go for a walk.

The Problem with Using AI for Goal Setting (And How to Fix It)

This is the problem with using AI for goal-setting. It's fantastic at generating possibilities but terrible at helping you choose. Ask it what matters most and it'll happily give you 15 priorities (which means you have zero priorities).

If you're a manager trying to set team goals, you've probably hit this same wall. Your company might not have adopted OKRs (or any real framework). You know goals matter, but the process either feels overwhelming or produces something uselessly vague like "improve team performance."

Here's what I figured out: The breakthrough wasn't getting AI to be more creative. It was getting it to help me be more disciplined.

Turning AI Chaos into Strategy: The 3-Level OKR Approach

My conversation with the LLM went through three phases:

Phase 1: "Tell me everything I should do" Result: Not great. A massive list that covered everything from revenue targets to social media strategy to operational systems. All important. None prioritized.

Phase 2: "Help me pick 2-3 things that matter most" Better, but still too abstract. "Grow revenue" and "build authority" sound good but tell you nothing about what to actually do.

Phase 3: Understanding Objectives, Key Results, and Actions This is where it clicked. I'm familiar with the OKR framework so used it to force specificity at three levels:

Objectives = where you're headed this year (2-3 max)
Key Results = how you'll know you got there this quarter (measurable)
Actions = what you'll actually do this month (concrete tasks you can check off)

The magic isn't the framework itself. It's that each level forces different decisions. Objectives force you to pick a direction. Key Results force you to define success. Actions force you to commit to doing specific things.

Real-World OKR Example: Building a Revenue Engine

Here's one of my actual objectives for Q1:

Objective 1: Build a Financially Resilient Revenue Engine Intent: Shift from "feast or famine" to predictable income

Key Results (Q1):

Secure £15,000 in revenue (60% weight)
Generate 10 qualified sales conversations via direct outreach (30% weight)
Lock in guaranteed base agreement with agency (10% weight)

Q1 Actions:

Jan: Finalize 2 pending SOWs; Clean CRM list; Launch Wave 1 outreach
Feb: Wave 2 outreach; Negotiate agency retainer floor
Mar: Win/Loss analysis; Pitch Q2 retainer upsell

Notice a few things here:

The Intent statement does real work. "Build revenue" could mean anything. "Shift from feast or famine to predictable income" tells you exactly what problem you're solving. When you're deciding how to spend your time in March, this guides you.

The weighting is ruthless. £15K gets 60% because it's the constraint (if I don't hit this number, I can't pay my bills). The other goals matter, but not as much. This isn't sophisticated math. It's admitting what actually keeps you up at night.

The monthly actions are specific enough to check off. "Finalize 2 pending SOWs" is binary (you either did it or you didn't). "Improve sales pipeline" would sit on the list forever because you'd never know if you were done.

Why This Matters for Managers

Most teams fail at goals because they're either too vague ("be more innovative") or too detailed (tracking 47 metrics no one looks at). The OKR structure forces you into a useful middle ground.

Annual Objectives give your team direction without locking you into tactics that might not work. You're saying "this is what matters for the next 12 months" without pretending you know exactly how to get there.

Quarterly Key Results let you adjust course without abandoning the plan. Every 90 days you ask: Are we making progress? What's working? What needs to change?

Monthly Actions keep the work tangible. Your team knows what to do this week, not just what they should achieve by December.

Here's another example from my goals that shows how weighting forces honest prioritization:

Objective 2: Validate the "Productized" Income Layer Intent: Launch a scalable cohort model without hourly work

Key Results (Q1):

Finalize Beta Curriculum MVP (20% weight)
Secure 10 Paid Seats for Beta Cohort (50% weight)
Achieve 8/10 Confidence Score from pilot feedback (30% weight)

Why is the curriculum only 20%? Because I can build curriculum in my sleep (it's not the constraint). The real risk is whether anyone will actually sign up (50%) and whether they'll find it valuable enough to recommend (30%).

This is the opposite of what most people would do. They'd spend 60% of their energy perfecting the content because it feels productive, then scramble to fill seats at the last minute.

Steps to Implement AI-Driven OKRs with Your Team

Here's the process:

1. Start messy, then constrain Let the LLM brainstorm everything your team could focus on. Don't filter yet. Then force it to prioritize: "Of these 15 things, which 3 would have the biggest impact for our team this year?"

2. Work top-down Get the annual objectives right first. Don't jump to tactics. If your objectives are "ship the new platform" and "reduce customer churn," every quarterly goal should ladder up to one of those.

3. Make it quarterly, not annual Trying to plan 12 months of actions is fantasy. Set the direction for the year, plan the quarter in detail, review every 90 days.

4. Test for specificity If you read an action item 8 months from now and wouldn't know what it meant, it's too vague. "Channel audit" tells you nothing. "Audit YouTube, LinkedIn, and email open rates to identify which channel drives most consulting leads" tells you exactly what to do.

5. Weight your Key Results honestly Don't give everything equal weight. Admit which results actually matter most based on your real constraints (budget deadlines, headcount freezes, whatever keeps your director up at night).

The Perfect AI System Prompt for Goal Setting

Here's what you can copy/paste to set up the conversation properly from the start:

I'm a [role] managing a team of [size] in [industry/function]. I need help setting team goals for the year using the OKR framework.

Here's what I need from you:
- Ask me questions about what matters most for my team this year (don't assume - ask)
- Help me identify 2-3 high-level objectives for the year with clear intent statements
- Break each objective into quarterly key results (measurable outcomes with weightings)
- For Q1 only, suggest 3-4 concrete monthly actions per objective
- Push back if I'm being too vague or setting too many priorities
- Format everything so it's easy to track in a spreadsheet

Before we start: I need you to be ruthless about prioritization. If I try to add a fourth objective, challenge me. If a Key Result isn't measurable, call it out. If an action is too vague, make me be specific.

Now, start by asking me about my current situation and what success looks like this year.

Download: Free Interactive OKR Template and Dashboard

For advanced users: I've created a JSON template and tracking dashboard you can customize. There's a step by step Readme guide included in the kit that's very easy to follow.

Download the OKR toolkit here

Note: I haven't had much time to test this so let me know any improvements you'd like me to fix/add.

Moving From Planning to Execution

The real value of this process isn't the framework (it's forcing yourself to make choices). AI will happily let you set 47 goals. Good management is about picking 3 and actually achieving them.

I score my OKRs at the end of each quarter, not weekly. Why? Because you need time to see if things are working. Weekly tracking of goals just creates busywork and anxiety. Do the work in January, check progress in February, adjust in March, score at quarter-end.

Use the LLM to think through options, but you have to do the hard work of deciding what matters. That's what the weighting forces you to do. That's what the Intent statement clarifies. That's what makes this actually useful instead of just another goal-setting exercise you'll ignore by March.

A Beginner's Guide to OKRs

If you haven't encountered OKRs before, it's worth reading Measure What Matters by John Doerr, who literally wrote the book on them. I've also created a short video using Notebook LM to explain the basics.

Watch the OKR explainer video