Home → Leadership
The Career Ladder: An Expectations Table, Not a Promotion
They asked in a one-on-one: “When do I become senior?” My answer was “we’ll see, you’re doing well”. Four months later they resigned. In the leaving conversation they repeated that sentence back to me, word for word.
- “We’ll see” is not an answer, it is a postponement. And the other person hears it as “I do not know”.
- Three axes are enough: scope, autonomy, impact. Technical skill is not separate; it shows up inside all three.
- A promotion is a recognition, not a reward. If someone already works at that level, the title is overdue.
- Two tracks, the same height. Management is not a higher level, it is a branch — and the move must go both ways.
- Seniority is not time served. “It has been five years” is not a reason, and five years of the same work is not progress.
- The table is not kept secret. An expectation nobody can read is not an expectation.
From the field: the answer that came four months later
I meant no harm in that conversation. They really were doing well and I really was thinking about it. But because I had no table, I had nothing concrete to say: where they stood on which axis, what was missing, when we would look again.
What they said on the way out was: “The problem was not that I did not become senior. It was that nobody could tell me what I needed to do.” They were right. They carried that question for four months while I had closed the conversation.
What I really missed is this: “when do I become senior” is not a question about a title. It is the question “do you see me, do you know where I am going?” And “we’ll see” answers it with a no.
Three axes
Most level tables swell into twelve-item lists that nobody reads. Three axes are enough:
- Scope. How large a piece of work do they own? A task, a feature, a system, or something that spans systems?
- Autonomy. How much direction do they need? Step by step, a goal to aim at, or do they find the problem themselves?
- Impact. Who does their work affect? Themselves, the team, several teams?
Why technical skill is not a separate axis: it cannot be observed on its own. “Writes good code” is not a level statement. “Without being asked, defined the boundaries of a system and produced an interface two teams now use” says all three at once.
| Scope | Autonomy | Impact | |
|---|---|---|---|
| Starting | Defined tasks | Regular direction and review | Their own work |
| Mid | A whole feature | Given a goal, finds the path | Part of the team |
| Senior | A system or an area | Finds the problem, moves in ambiguity | The whole team; makes others faster |
| Advanced | Cross-team concerns | Also decides what will not be done | Several teams; decisions leave lasting traces |
Nobody is at the same level on all three axes at once. The table is not there to say “you are here”; it is there to make the conversation concrete. “On scope you are senior; on impact you are in the middle: nothing you did in the last six months made anyone else faster” is far more useful than “we will see” — especially when it is bad news.
A promotion is a recognition
The most common mistake is running the order backwards: giving the title first and waiting for the person to grow into it. It is a well-meant mistake, usually made to keep someone from leaving. It usually hurts both sides.
- The person starts doing the next level’s work
- It is seen consistently for a few months
- The title follows — it records what is already true
Nobody is surprised on promotion day, because everyone could already see it.
- The title is given (“for motivation”)
- The person is left with an expectation they have not met
- Six months later: “you are not delivering at your level”
The unfair part: they never made that promise. You did.
A natural consequence: the promotion decision is not made at promotion time. It starts months earlier, with “here is the work that will take you to that level”. That work is a real responsibility that is delegated, not an exercise.
Two tracks, the same height
If you make management the next rung of the ladder, you push good engineers into becoming bad managers. There should be two tracks, and the same levels should be worth the same:
TECHNICAL TRACK MANAGEMENT TRACK
------------------------------------------------------------
Senior engineer <-----> Team lead
Staff engineer <-----> Senior manager
RULE 1 : Same row, same level. Different title, same value.
RULE 2 : The move goes BOTH WAYS. Returning from management to
the technical track is a change of preference,
not a step backwards.
RULE 3 : Rule 2 does not work unless it is written down. People
assume "if I go back I will be seen as having failed"
and never try.I learned rule three the hard way. Two people tried management; one was unhappy and said nothing for a year, because they did not know how a return would be received. An unwritten rule is read in its most cautious interpretation. Why management is a different job rather than a higher level is in a separate post; if your ladder does not reflect that, it produces the wrong incentive by itself.
Seniority is not time served
Two sentences, both wrong and both common:
- “It has been five years, they should be senior by now.” Time is not an axis. Five years at the same scope can be the first year repeated five times.
- “They write the best code, so they are the senior.” Individual output is only a part of the impact axis. Output that makes nobody faster and leaves nobody any context does not count as senior in the table.
The second is especially sneaky, because it is easy to reward and easy to see. Promote the fastest coder and you tell everyone on the team that producing a lot matters more than sharing what you know — and six months later nobody has time for code review.
How I run the conversation
Once a quarter, inside the one-on-one, 20 minutes. Three questions:
- “Where do you see yourself in the table?” They speak first. Most people place themselves lower than they are, and that alone is worth discussing.
- “Here is where I see you, and why.” With concrete examples. Without examples it is not an assessment, it is an impression.
- “The next level’s work looks like this; this quarter I am giving you one of those.” Without concrete work the conversation becomes a wish.
Keeping this inside the one-on-one matters: when we set up a separate “career meeting”, it became a yearly ceremony and nobody asked anything in between.
What to track
| What | Why |
|---|---|
| How often “when do I get promoted” is asked | If it is rising, the table either does not exist or is not read |
| People surprised by their own promotion | Should be zero; surprise proves the conversation never happened |
| Seniors staying on the technical track | If everyone drifts to management, the tracks are not the same height |
| Moves back from management to technical | Zero is not good news; it means the move is one-way |
What did not work for me
- A detailed level document. Seven levels, twelve bullet points each. Nobody read it, including me. Cutting it to one page with three axes is what got it discussed.
- Keeping the table private. We said “let us settle it among managers first” and it stayed private for six months. During that time nothing about expectations changed, because nobody knew what they were. An expectation nobody can read is not an expectation.
- Tying promotion to the performance review. They answer different questions: performance is “how did this period go”, level is “how big is the work they do”. Merge them and everyone who has a good quarter expects a promotion — another version of the confusion I described in the performance post.
Checklist
- Can everyone on the team see the table?
- Does everyone know what level I see them at — and why?
- Is “the next level’s work” defined as concrete work for each person?
- Was anyone surprised by the last promotion?
- Are the technical and management tracks worth the same at the same level?
- Is it written down that moving back is not a failure?
- Have I made any decision that confused seniority with time served or raw output?
- When did I last have this conversation?
Conclusion
When I lost that person I did not have a bad table — I had no table at all. What I did over the next six months was not complicated: one page, three axes, four levels, visible to everyone. It had no magical effect; nobody read it and felt happy.
The only thing that changed was what the conversations were about. “When do I become senior” turned into “what do I need to do on the impact axis”. The second question has an answer. The first one did not.
The test: can everyone on your team say, in one sentence, what they need to do to reach the next level? If not, you have a ladder but nobody can see the rungs.