Designing Your Framework
For Project Leads.
A lot of people can help build frameworks, so agreeing on a few parameters up front keeps content consistent, makes the build easier, and makes positions easy to compare across teams. Once they're agreed upon, put them into a template so every team starts the same way.
1. Teams and tracks
Each team (Engineering, Marketing, Finance and so on) has its own framework. Tracks split a team into progression paths. Most teams have Individual Contributor (IC) and Management tracks; larger teams may split by discipline. A team with only two or three positions might need just one track.
List your teams.
Decide whether you'll add standard IC and Management tracks, or let HRBPs and Function Leads create discipline tracks.
Assign an HRBP and/or Function Lead to each team.
2. Positions
Pick one naming approach and use it everywhere:
Approach | Example | Pros | Cons |
|---|---|---|---|
Explicit | Senior Frontend Engineer | Very clear expectations | Slower to build; more skills and tracks to manage |
Abstract | Mid-level Software Engineer | Fewer positions and skills; faster rollout | Less nuance for each job |
Numbered | SWE1, SWE2 | Works when job titles are inconsistent; no retitling | Not always clear which titles map to which position |
If you need a new levelling structure, sort that out before you build. Also decide whether positions will show salary ranges. They're optional. Depending on your plan, you can also split them by location.
A good position has:
A clear title that follows your naming convention.
A short description that sums up what's expected at a glance. Base it on the job description, but keep only the key points. The AI Assistant can draft one for you.
Required skills, with the level needed for each. Use a mix of domain-specific, core and leadership skills.
A label, if you use them: a short code showing track and seniority. For example, DM5 could mean level 5 in the Design Management track.
3. Skill categories
Every skill sits in a category. Keep to three or four, with clear definitions, and use them in every framework. A common set:
Domain-specific: technical skills such as Copywriting or Data Analysis.
Core: transferable skills such as Communication. Values-based skills often go here.
Leadership: such as Giving Feedback, Developing Others and Hiring.
4. Number of skills
We recommend 8 to 12 skills per position. Too many overwhelms people, makes conversations less focused and makes quality harder to maintain. Fewer skills also means a faster rollout, and you can add more later. IC positions usually lean towards domain-specific skills, and leadership skills grow with seniority.
5. Core skills
Core skills are the skills everyone needs, whatever their team or position. They keep frameworks consistent and make it easier to move between teams.
Your company values are a good starting point. They're familiar, they link values to everyday work, and they feel more relevant than generic skills. If you have more values than skill slots, build values into other skills. For example, a "Thinking outside the box" value could become an example in Communication: "You come up with new ways to share information."
Core skills can feel too generic for some teams. Instead of creating lots of near-duplicate skills, an admin can create the main skill and editors can create variants tailored to their team. Variants depend on your plan.
6. Put it into a template
Build one team the way you want every team to look, with your tracks, naming convention, categories and core skills. Lock what shouldn't change, then convert it to a template so other teams start from it. See Templates and locked skills.
Tip
If you're using the AI Framework Builder to draft frameworks, agree these parameters first so you can check the drafts against them.
