Unpopular opinion: OKRs don't cause problems, they surface them.
If you have a clear strategy, and everyone understands the company's goals and their role, OKRs should write themselves.
If not...focus on the root cause.
Here are 6 root causes of OKR migraines:
Peanut butter (speading your attention across too many things) - Assume that 90% of your results will come from 10% of the stuff you try. Focus on finding your big lever (if you haven’t yet), not pulling the small ones even harder.
No prizes for effort - The "R" in "OKR" stands for results not effort. In lead generation, for example, you would focus on the number of qualified leads not work units (i.e. number of calls and emails). Hold your team accountable to the results and let them figure out the best way to get there. “We don't hire smart people and tell them what to do. We hire smart people so they can tell us what to do.” - Steve Jobs
Herding Cats - If teams are pulling in different directions and fighting for resources, ask each of them to connect their main KRs to your North Star metric (e.g. monthly active users). When everyone's solving for the same outcome, it's easier to prioritise work across different teams.
ADHD - Founders have ADHD and that’s a feature, not a bug. But teams can’t execute without sustained focus. When new ideas come, don’t abandon your strategy. Test them instead. Identify your risky assumptions and run experiments. Don't change directions when you get an idea, change directions when you validate it.
JFDI or DFTU? For some projects, speed trumps quality, so JFDI (just f*ing do it.) But others, like security and uptime, are mission critical, or DFTU ("don't f**k this up.") Those are mutually exclusive, so make sure everyone is on the same page from the start. Literally go through each goal and agree on the expected level of quality vs. speed.
Sandbagging - OKRs shouldn't be easy. Include at least one "stretch" goal that challenges you to grow, learn something new or reach a new level. This doesn't necesarrily mean fixing a weakness. Encourage strong performers to build on their strengths.
Simple next step
When intelligent people disagree they’re usually working from different assumptions. So use this list to unpack your assumptions and get to the root cause - don't blame the framework.