Most companies build competency frameworks backwards. They start with generic buzzwords — "leadership," "communication," "strategic thinking" — and slap a 1-5 rating scale on them. Then wonder why nobody can actually use the thing during performance reviews.

Here's the fix I've used with dozens of teams: build the framework from observable behaviors, not traits.

Instead of rating someone on "leadership," define what leadership actually looks like at each level in your org:

• Level 1: Leads a single project, delegates clear tasks, follows up without micromanaging

• Level 2: Leads a small team, resolves conflict without escalation, coaches direct reports

• Level 3: Leads leaders, sets team-level strategy, is accountable for outcomes beyond their own work

The moment you make it behavioral, three things happen:

1. Managers can actually give specific feedback ("you're not yet delegating clearly" vs. "be more of a leader")

2. Employees know exactly what to do to level up — no guessing

3. Promotion decisions stop being political because there's a shared, observable bar

The hardest part isn't writing the framework — it's getting managers to rate consistently against it. That's where most rollouts die. If you're building one for your org, happy to share what's worked (and what hasn't) — drop a comment.