Employee Retention

Retention is usually measured as the inverse of attrition, which imports the same problem: it treats every departure as a loss and every stayer as a success.

The useful framing is narrower. Retention is keeping the people whose leaving would hurt, for long enough to get the value of having hired them. That makes it a targeted activity rather than a general one.

Two consequences follow.

• Retention effort should be differentiated. Organisation-wide initiatives spend most of their budget on people who were not going to leave and would not be missed if they did.
• Some retention is undesirable. People who have outgrown a role and cannot move, or who are disengaged but comfortable, are being retained by inertia, and that is a cost rather than an achievement.

The measure that reflects this is regretted attrition, and the tenure at which it occurs. An organisation losing few people but losing its best ones at the two-year mark has a retention problem that its headline figures conceal.

The pattern in the right-hand column is the point. Retention programmes concentrate on the lever with the least effect, because it is the one HR can implement without changing how managers manage or how pay is set.

Pay compression deserves particular attention because it is both consequential and quietly created. Paying market rates to new hires while giving existing staff percentage increments guarantees that longer-serving people fall behind. They do not usually negotiate when they discover it; they look.

By the time someone resigns, they have a signed offer and a plan. The window that mattered closed months earlier.

• Ask managers directly and regularly who on their team they would be concerned to lose, and what would keep them. Managers usually know, and are rarely asked in a way that requires an answer.
• Watch tenure clusters. Exits concentrate at predictable points, commonly just after appraisal outcomes, after bonus payment, and around the two to three year mark.
• Read engagement data by team rather than organisation-wide, since the organisation-wide average conceals exactly the concentration that matters.
• Track internal mobility. Where people cannot move sideways, they move out. A low internal transfer rate alongside rising attrition is a strong signal.
• Review pay against current market entry rates for existing staff at each grade, before rather than after someone compares notes.
• Talk to people who considered leaving and stayed. They are the most informative population and almost nobody asks them.

A counter-offer addresses the last item on a list of reasons, at the point where the person has already gone through the effort of interviewing elsewhere.

Three reasons they disappoint.

• Money was rarely the whole reason, and the other reasons remain unchanged the following week.
• The relationship changes. The employee is now known to have been looking, and both sides behave differently afterwards.
• It teaches the organisation's lesson to everyone watching: the way to get a raise here is to get an offer elsewhere. That cost is borne by people who never resign.

Where a counter-offer is genuinely worth making, and occasionally it is, the honest version fixes the actual cause rather than only the number, and it is accompanied by an explicit acknowledgement that the organisation should have acted earlier.

The better use of the same money is almost always to correct compression proactively across the affected group. It costs more in total, it does not reward resignation, and it prevents the resignations that have not happened yet.