The best-supported figure is Gallup's estimate that replacing an employee costs one-half to two times their annual salary. A widely quoted survey found workers believe 42% of their job knowledge is unique to them. Both are useful, but neither measures knowledge loss directly. The most reliable number is one you calculate from your own leavers.
Statistics about knowledge loss get repeated until nobody remembers where they came from. Here is what the main ones actually say.
The numbers that hold up
| Figure | Source | What it measures | Caveat |
|---|---|---|---|
| Replacement costs 0.5x to 2x annual salary | Gallup, 2019 | Total cost of replacing an employee | A broad estimate; includes hiring, ramp-up, lost output |
| About 200% (leaders), 80% (technical), 40% (frontline) | Gallup | Replacement cost by role type | Estimates from a review of studies |
| 42% of job knowledge is unique to the individual | Panopto/YouGov, 2018 | Workers' own estimate | Self-reported; vendor-commissioned; US only |
| 5.3 hours a week waiting for help or information from colleagues | Same Panopto survey | Self-reported time lost | Same caveats |
The numbers to question
Some widely shared figures are hard to trace to a primary source, or come from reports that are now old. Before you quote a statistic:
- Find the original. Not a blog quoting a blog.
- Check the date. Workplaces have changed a lot since 2018.
- Check who paid. Vendor surveys are not worthless, but say so.
- Check what was measured. Self-reported estimates are not measurements.
How to estimate your own cost
A simple model:
Annual knowledge loss cost ≈ leavers per year × hours lost per leaver × loaded hourly cost
Where "hours lost per leaver" includes:
| Component | How to estimate |
|---|---|
| Successor ramp-up beyond normal | Ask recent successors how long until they were fully productive, minus what you expect for any new hire |
| Colleagues answering questions | Hours per week spent helping × weeks |
| Rework and mistakes | Incidents traced to missing knowledge in the past year |
| Re-made decisions | Meetings that reopened settled questions |
Example (illustrative only): 30 leavers a year × 120 hours × ₹1,500 an hour ≈ ₹54 lakh. Change the inputs to your own figures. Our memory debt calculator runs a version of this.
What to do with the number
- Compare it with the cost of better handovers (a few hours per leaver).
- Use it to prioritize roles with key-person risk.
- Track it each year as you improve your process.
Where Reattend fits
We built Reattend because the cost is real even when the exact number is uncertain. It keeps decisions, commitments and context as work happens, so a successor can ask instead of guessing. We would rather you measure the problem with your own data than take any single statistic, ours included, on trust.
Frequently asked questions
How much does it cost when an employee leaves?
Gallup estimates the cost of replacing an employee at between one-half and two times their annual salary, and later broke this down as roughly 200% for leaders and managers, 80% for technical professionals and 40% for frontline employees.
How much of an employee's knowledge is unique?
In a 2018 YouGov survey of 1,001 US workers commissioned by Panopto, respondents estimated that 42% of their job knowledge was unique to them. It is self-reported and vendor-commissioned, so treat it as indicative.
How do I estimate knowledge loss cost for my organization?
Multiply the number of leavers per year by an estimate of time lost to rebuild their knowledge (successor ramp-up, colleagues' time answering questions, rework and mistakes) and by a loaded hourly cost. Use your own data where possible.