Senior engineer attrition: see it coming while it can still be changed
Turnover research puts the cost of a resignation at up to twice a year’s salary, and that is before the knowledge they take with them. The resignation is the last event in a long chain of visible signals. What follows is how to read that chain early; the signals list and cost worksheet at the end are free.
Attrition is a lagging indicator of team conditions
By the time retention reaches the leadership agenda, someone has resigned, and the response happens at the most expensive possible moment: counteroffers, backfills, knowledge-transfer scrambles. But senior engineers disengage from the future before they leave the present, and every step of that disengagement is observable in team behavior: speaking up less, proposing nothing, working alone, learning nothing new.
None of this is folk wisdom. Replacement research puts total turnover cost at 50 to 200 percent of annual salary depending on seniority, and Visier’s workforce data found coworkers 9.1 percent more likely to quit in the months after a teammate leaves, an effect that roughly doubles in small teams or when the leaver was a high performer.
The AI era added new drivers to the old list. Engineers spend their days with agents instead of teammates, and even Anthropic’s own Claude Code team found the work turning lonely. Growth stalls while output climbs, which is skill atrophy from the inside. And when a senior does leave an AI-heavy codebase, they take the only full mental model of what the agents built, turning one resignation into instant comprehension debt.
How Aurora Coach addresses it
What you get is fewer surprise resignations: the conditions that push seniors out become visible as trends while they are still fixable, and fixing them is owned by the team rather than discovered in exit interviews. Aurora Coach runs that as a continuous improvement loop, Sense, Analyze, Recommend, Refine, Commit, Execute, Re-evaluate. Retention lives mainly in the Foundation domain (growth, autonomy, purpose, psychological safety), with Alignment (ownership, impediments) alongside.
- Sense + Analyze Each period the whole team answers, one by one: energy, growth, connection, friction. Voice drops, isolation drift, and stalled learning register as trends in exactly the signals above, while they are still conversations rather than resignations.
- Recommend + Refine + Commit Trends turn into proposals: a growth commitment, a connection ritual, ownership spread across more heads, each with rationale and success criteria. The team votes and the lead refines; deciding stays human. Retention becomes team practice, not a counteroffer scramble.
- Execute + Re-evaluate Commitments run alongside delivery, and the following period shows whether the signals moved: people speaking up again, learning restarted, knowledge flowing. The conditions that make seniors stay are the same ones that make teams good, so the work pays twice.
The signals are the same ones three sibling pages work in depth: developer loneliness for the isolation drift, AI skill atrophy for stalled growth, and psychological safety for whether problems get said at all.
What does it cost to lose a senior engineer?
Turnover research puts the total cost at 50 to 200 percent of annual salary, rising with seniority, and that is before the months a successor needs to reach full productivity. The larger costs are less visible: comprehension debt on every system only they understood, extra load on the people who stayed, and the way one senior departure re-prices everyone else’s decision to stay.
What are the early warning signs an engineer is about to leave?
They disengage from the future before they leave the present: speaking up less in retros, no longer proposing improvements, working increasingly alone, stalled learning, dropping discretionary work like mentoring and docs, and knowledge that stops flowing outward. Each is observable in team behavior months before a resignation letter.
Can you improve retention without counteroffers?
Counteroffers act at the resignation, which is the most expensive and least effective moment. The conditions that make senior engineers stay are upstream and team-shaped: growth that continues, work with real ownership, connection to the people around them, and problems that get fixed when raised. Those are improvable conditions, and improving them is cheaper than replacing anyone.
The six early-warning signals, free
Read them against each senior on your team. One signal is a mood; three or more, sustained for a month, is a trajectory.
- Stopped speaking up in retros and check-ins The person who used to argue about how the team works stops arguing. Caring about the process ends before caring about the job does.
- Stopped proposing improvements Suggestions are investments in a future here. When someone stops making them, they have often stopped picturing that future.
- Working increasingly alone Pairing stops, reviews shrink to rubber stamps, agents replace teammates as the first person asked. Isolation precedes exits, and AI-heavy workflows accelerate it.
- Growth has stalled Nothing new learned in months, no task that stretched them. Senior engineers rarely leave over money alone; they leave when staying stops making them better.
- Only critical-path work The discretionary contributions go first: mentoring, docs, tooling for others. A senior doing exactly their tickets and nothing else is already halfway out.
- Knowledge stopped flowing outward They used to teach and write things down; now knowledge accumulates with them and stays. Watch this one, because it also prices the exit.
The cost-of-one-exit worksheet
Five lines that turn a resignation into a number your leadership already cares about:
- Replacement cost Turnover research puts the total at 50 to 200 percent of annual salary, rising with seniority. Write your number down.
- Ramp time The months before a replacement reaches the productivity of the person who left, in your codebase, with your context. Estimate it honestly; it is rarely a quarter.
- Comprehension walking out Every system only they fully understood is now comprehension debt. In AI-heavy codebases the person who supervised the agents may be the only one who ever understood the result.
- Team drag Interim coverage, extra review load, and on-call gaps land on the people who stayed, which is how one resignation starts pricing the next.
- The contagion term Attrition clusters. The first senior departure re-prices everyone else’s decision to stay. Whatever number you wrote above, it is the optimistic case.
Not ready to change anything today? You already have the signals list and cost worksheet above, copy buttons and all. If you want one improvement loop like these in your inbox each month, leave your email.
The signals and the worksheet are free to use tonight. The trial adds the every-period sensing that turns them into trends. Map your ROI below turns a few questions about your team into the numbers leadership asks for, no signup, a few minutes.