Most career ladder templates are built for companies with hundreds of engineers. At ten people, a four-level ladder with clear, observable behaviors beats a twelve-level framework nobody reads. Simple and used beats comprehensive and ignored.

A ladder that fits a small team

Level	Scope	What “good” looks like
Engineer I	Well-defined tasks	Ships reliably with guidance; asks good questions
Engineer II	Small features end-to-end	Owns a feature from spec to production without hand-holding
Senior Engineer	Cross-cutting problems	Makes architectural calls; unblocks others; mentors
Staff Engineer	Org-wide leverage	Work multiplies through others, not just their own output

Four levels, each with one sentence of scope and one sentence of what good looks like. That’s enough for a company under thirty engineers — add complexity only when you actually have people it fails to describe.

Write down what changes between levels, not just titles

The jump that confuses people most is Senior to Staff, because the title implies “more senior IC work” when it should mean “more leverage through others.” Be explicit: a Staff engineer’s job is measured partly by what the team around them ships, not only by their own commits. If you don’t say this out loud, people optimize for the wrong thing on the way there.

Promotion criteria, written before the promotion conversation

Decide the criteria before someone is up for promotion, not while deciding whether to promote them. Deciding in the moment is how bias creeps in, rewarding whoever’s most visible rather than whoever’s actually operating at the next level. For each level, write two or three observable behaviors, not vague qualities:

• Not “shows leadership.”
• Instead: “has mentored at least one engineer through their first two production incidents” or “has made an architecture decision that another team adopted.”

Tie it to compensation transparently

State the salary band for each level, even roughly, and update it when the ladder changes. A ladder without a comp connection reads as a manager’s opinion of you dressed up as a framework. A ladder tied to a band gives people something to actually work toward, and gives you a defensible answer when someone asks why a peer got a bigger raise.

Review it twice a year, not never

Ladders rot. A level that made sense for a five-person team stops fitting at twenty. Set a recurring reminder to revisit it, even a 30-minute review, rather than letting it calcify into something everyone quietly ignores because it stopped matching reality two years ago.

The point of doing this at all

A ladder isn’t bureaucracy for its own sake, it’s the thing that lets you have a promotion conversation that’s about observable behavior instead of gut feel and who you like talking to. Skip it and you’ll end up promoting based on recency and charisma, which is exactly the outcome a ladder exists to prevent.

---