A Career Development Plan Is Goals Plus an Evidence Loop

I have written the same career development plan three Januaries in a row. Same goals, same optimism, same November rediscovery of the document, untouched since week two. The fourth version survived, and not because I became more disciplined. I made the plan smaller and gave every goal a place where proof could pile up without me thinking about it. That is the whole difference between a plan and a document: a document gets filed, a loop keeps running. This is the structure that survived, with the one-page template.

What a career development plan actually is

Strip the ceremony and a working plan has three parts: a short list of goals, an evidence loop that feeds each one, and a review cadence that forces an honest look at the gap between them. That is the whole machine. Goals give the record a direction. Evidence tells you whether you are moving. The cadence is what keeps either from going stale.

Just as important is what a plan is not. It is not a five-year ladder fantasy; nobody can see five years out, and pretending to produces goals too vague to act on. It is not the form HR asks for in onboarding week. And it is not a list of courses. Courses are sometimes evidence of progress. They are never the goal.

Why most development plans die by March

Three failure modes account for nearly every abandoned plan. First, vague goals. Decades of goal-setting research are unambiguous: specific, difficult goals produce higher performance than 'do your best' intentions, because a vague goal gives you nothing to check yourself against. 'Get better at communication' cannot fail, so it cannot succeed either.

Second, no evidence loop. A plan that lives in a drawer has no input stream. Nothing accumulates against the goals, so revisiting the plan feels like an exam you forgot to study for, and you stop revisiting it. Third, annual cadence. Companies from Adobe to Deloitte abandoned the once-a-year cycle for exactly this reason: feedback that arrives months after the work is too late to steer anything. Your private plan fails on annual cadence for the same reason their reviews did.

Step 1: Pick three or four goals, one per category

Four categories cover what a career actually grows along. Skill: something you can do at the end of the year that you cannot do now. Scope: a bigger problem trusted to you. Visibility: the right people knowing what your work changed. Optionality: choices you have that you did not have before, inside or outside the company.

One goal per category, four maximum. Past four, goals stop getting evidence streams and the plan quietly reverts to a wish list. If everything feels important, the categories force the argument with yourself that a blank page never does.

Step 2: Wire each goal to an evidence loop

This is the step every template skips, and it is the one that decides whether the plan survives. For each goal, name where its evidence will come from before the year starts. The unit of progress is not a feeling of momentum. It is an entry: dated, specific, tagged to the goal it serves.

• Skill goal — entries logging reps: the migration you led, the talk you gave, the review feedback that named the skill.
• Scope goal — entries marking trust events: the project handed to you, the decision made without sign-off, the incident you owned.
• Visibility goal — kept praise: forwarded emails, screenshots of kudos, the meeting where your work was cited.
• Optionality goal — external signals: recruiter outreach, referral requests, an interview taken to calibrate.

The capture habit can be tiny. Three lines a day, or even just forwarded emails and shared screenshots on the days you will not write. What matters is that every goal has a stream, because a goal with no stream is invisible by June.

Step 3: Set the review cadence

Two sessions, both small. Weekly, five minutes: scan the week's entries and tag any that serve a goal. Quarterly, thirty minutes: for each goal, count the evidence, score progress one to five, and decide whether the goal earns another quarter. The quarterly session is the plan; everything else is feeding it.

Do not wait for your manager to run this for you. Gallup's workplace data has said for years that structured development conversations are rare, and LinkedIn's learning research keeps finding the same gap: people rate career development among the top reasons to stay, and most get no organized version of it. The cadence you run yourself is usually the only cadence there is.

The one-page plan template

Five fields per goal. One page covers the whole plan.

• Goal — one sentence, testable verb, this year.
• Why now — the one-line reason this goal earns a slot this year and not next.
• Evidence stream — where dated proof will accumulate: entries, forwarded praise, shipped links.
• Quarterly checkpoint — the question you will score one to five each quarter.
• Kill condition — what would make you drop or rewrite this goal without guilt.

Worked example, senior engineer aiming at staff scope: Goal — lead a cross-team technical decision end to end this year. Why now — last review named influence beyond the team as the gap. Evidence stream — entries on design reviews led, RFCs authored, the moment another team adopted the proposal. Quarterly checkpoint — did my technical judgment shape work outside my team this quarter? Kill condition — re-org removes cross-team surface for two consecutive quarters; rewrite the goal around the new org.

Feeding the plan into review season

Run the loop and review season changes shape. The self-review stops being archaeology, because the evidence is already sorted by goal. The development conversation stops being improvised, because the quarterly scores are the agenda. Do not rebuild anything in November; the plan was the source document all along.

When to change the plan

A plan you never revise is a plan you stopped believing. Kill conditions exist so that dropping a goal is a decision, not a slow fade. When a goal dies, write one line on why, pick its replacement deliberately, and move on. A dropped goal with a written reason is a healthy plan. Four goals silently ignored since February is the failure mode you were avoiding.

You do not need January to start. Pick one goal tonight, give it an evidence stream, and write the first entry against it. The other two or three goals can arrive this quarter. The plan fits on one page and the loop costs minutes a week. A plan that starts small and runs beats every beautiful document that never made it past week two.