The Dead Sea Effect: Why Your Best People Evaporate First and the Salt Stays Behind

60-Second Summary

Bruce F. Webster (2008): the best engineers have the highest option value on the market, so they leave first when conditions deteriorate. The least mobile stay — and their concentration rises over time.
This is a distillation effect, not a bad-hire effect. You can hire perfectly and still end up with a salt-concentrated team five years later if you don't attend to the mechanism.
The dangerous signature: quiet, tenured, low-conflict, low-output, hostile to change, expert at process, unrecruitable. Every 'we've been trying to fix this team for years' team.
Individual performance management barely touches it because the salt is legally competent, procedurally correct, and socially embedded. Structural intervention is the only lever.
Fixes: cap tenure-based veto power, force role rotation, protect the freshwater (top-quartile mobility risk), rewrite the promotion criteria that reward tenure over impact, and — occasionally — dissolve and re-form the team.

Every engineering VP has met this team. Twelve people, average tenure eight years, no attrition, no obvious low performers, and yet nothing ships. Reviews are respectful. Retros are procedural. New leaders arrive with mandates, spend eighteen months trying to change things, and leave. The team's composition barely moves. This is not a management failure — or rather, it is, but not the one you think. It's the Dead Sea Effect, and once you see the mechanism you can never un-see it.

What Webster actually observed

“The reality is that most large IT shops end up like the Dead Sea. The more skilled and effective IT engineers are the ones most likely to leave — they have the highest 'evaporation rate.' What remains is 'the residue' — the least talented and effective IT engineers.”
— Bruce F. Webster, 'The Wetware Crisis: the Dead Sea Effect' (2008)

Webster was a long-time IT consultant who noticed the same pattern across dozens of enterprise engagements: the organisations most in need of improvement had, on paper, no attrition problem. They had the opposite — 'stable' teams whose stability was itself the disease. He named it after the Dead Sea because the mechanism is identical: freshwater flows in, water evaporates off the top, and what stays behind gets saltier every year. The Dead Sea is now 34% salt — nearly ten times ocean concentration — because the mechanism is one-directional and slow.

Webster's original piece was published on his personal blog in April 2008 and circulated for a decade in engineering-management circles before mainstream HR encountered it. It has never appeared in a Gartner report. Most CHROs have never heard of it. This is exactly why it is useful — it names something HR's own vocabulary systematically hides.

The distillation mechanism, in detail

Why the best leave first — the option-value argument

1
Higher external option value
Top-quartile engineers have 5–10× the recruitment inbound of median performers. When conditions deteriorate, they exercise the option first. Median performers can't leave as easily — their market is thinner.
2
Higher opportunity cost of staying
For a strong engineer, staying in a low-output team means their compounding curve flattens. They know it. The internal cost of staying is real to them and invisible to management.
3
Lower switching cost
They have references, portfolio, network, savings. The bar to leave is low. For long-tenured median performers, all four are absent — leaving is genuinely hard.
4
Higher sensitivity to slope, not level
Strong performers detect trajectory. They leave when the second derivative turns negative — long before the org-wide signals fire. By the time engagement scores drop, they're already gone.
5
Selection into the referral network
The freshwater takes the freshwater with them. One senior IC leaving typically pulls 2–3 more within 18 months (Kramer & Levitt, on developer mobility clusters). Salt does not have a cluster.

3–5×

productivity spread between top and bottom quartile knowledge workers

Aguinis & O'Boyle (2012); the top quartile is what evaporates

60–70%

of voluntary tech attrition is regretted (top-half performers leaving)

LinkedIn Talent Insights, various years

8–12 yrs

average time for a team to visibly 'salt out' under weak leadership

Webster (2008); confirmed anecdotally across enterprise engagements

How to recognise a salt-concentrated team

Average tenure > 6 years with near-zero recent hires that stuck.
Every meeting references history the newcomer can't access ('we tried that in 2019').
Procedural competence is high. Output is low. Nobody is 'failing' by any measurable standard.
New hires last 12–18 months and leave describing the team as 'political' or 'immovable'.
The team's own retros produce process-improvement actions that never change output.
Requests for headcount growth are resisted by the team itself — 'we don't need more people, we need to protect what we have'.
The strongest ICs on paper are the least visible in delivery.

The performance-management trap

The salt is not underperforming in any way HR can act on. They are on time, procedurally correct, non-conflictual, and embedded in tribal knowledge. Trying to PIP them fails legally and morally. This is why performance management is the wrong instrument. The Dead Sea Effect is an org-design problem, not a performance-management problem.

Structural interventions that actually work

Six moves that shift a Dead Sea team

1
Protect the freshwater actively
Identify top-quartile mobility risk and negotiate real conditions to keep them: scope changes, autonomy, comp, exit from the affected team. Losing them extends the disease by another five years.
2
Cap tenure-based veto power
The salt's power comes from tribal knowledge that is treated as veto right. Rewrite RFC processes so decision rights are role-based, not history-based.
3
Force role rotation
Every senior IC rotates project or system ownership every 24–36 months. Breaks the moat and rebuilds transferable skill. Some will refuse and leave. That is the point.
4
Change the promotion criteria
Audit your ladder: if 'depth of system knowledge' and 'tenure in domain' are heavily weighted, you are actively selecting for salt. Reweight toward observable impact and cross-system contribution.
5
Insert 3–5 senior hires simultaneously
One senior hire gets absorbed and neutered (see Absorptive Capacity). Three or more form a critical mass that can shift norms. Never solve a Dead Sea team with a single hire.
6
Dissolve and re-form
The nuclear option: reorganise the team, redistribute people across other teams, and rebuild with a deliberately mixed seed. Sometimes the salt is not shift-able and the only intervention is to end the container.

What the honest board conversation sounds like

'This team has been the same twelve people for six years. Two of them are structural. Two are strong ICs we haven't lost yet. Eight are procedurally competent and immovable. The plan is: promote and protect the two strong ICs, redistribute the eight over the next 18 months, and rebuild the team with a mixed seed of six hires. This is not a performance issue. It is a chemistry issue and the only intervention is to change the container.'

FAQ

Frequently asked questions

Isn't this just a euphemism for firing older employees?

No, and this is where the analysis has to be careful. Salt correlates with tenure, not age. Long-tenured people who kept growing are freshwater, not salt. The signature is stagnation, not seniority. The intervention is role change and structural reset, not termination — because the salt is almost never legally terminable.

Doesn't this contradict psychological-safety research?

No. Psychological safety enables strong performance in mobile teams. In an immobile salt-concentrated team, psychological safety becomes protection for stagnation — 'don't challenge us, we're safe here'. Safety is necessary but not sufficient.

How does this interact with the Peter Principle?

The Peter Principle explains why individuals rise to their level of incompetence. The Dead Sea Effect explains why the team around them concentrates over time. Together they explain the modal large-company engineering organisation.

What about remote teams — same effect?

Worse, actually. Remote teams have lower ambient signal, so the trajectory-detection freshwater relies on is weaker. They tend to evaporate later, but the salt concentrates faster once the pattern starts.

Takeaways

The Dead Sea Effect is a distillation, not a hiring failure. You can hire perfectly and still end up salt-concentrated.
The best leave first because they have the most options and the sharpest trajectory-detection.
The salt is legally competent, procedurally correct, and immovable by performance management.
Only structural interventions work: rotate roles, cap tenure-based veto, protect freshwater, hire seniors in clusters, and — sometimes — dissolve the container.
If you've been trying to fix a team for five years and it hasn't moved, you are not underperforming. You are running the wrong instrument.