Targeted Advisory case study

Who to Keep? Rate the Post and the Person Separately

280positions and their holders reviewed
2levels rated separately: the post and the person
6weeks to rate every division, over the year-end break
100%of posts in scope rated and calibrated

The situation

The question behind the question.

Funding was tightening, and this organisation was restructuring across all its divisions. Hundreds of fixed-term contracts were due to end within the year, with roughly 1 extension slot for every 3 of them. At the same time, long-serving staff could apply to leave early. Without a shared definition of who was critical, every division would argue for its own.

As it was asked

“Tell us which contracts we can afford to extend.”

As we reframed it

Define who is critical, the post and the person separately, rate every role on one scale, and test the rule before any name is attached, so one review decides extensions, exits and retention.

How we worked

How the work unfolded.

01

Separate the post from the person

We started by splitting one question into 2. Every post fell into one of 4 types, critical, specialised, core or flexible, and was rated on its impact on strategy and business continuity and on how hard it would be to replace. Its holder was rated on strategic contribution and scarcity of expertise. Each criterion was scored 1 to 5 against written anchors, so a 4 meant the same to every rater. 2 more fields asked how long the post would stay critical, and how soon its holder might leave.

02

Rate where the knowledge is

Then the ratings, made by those who know the work. Division heads and their HR business partners first got a briefing pack: definitions, worked cases, answers to the questions raters ask. Each division head then rated their own posts and people in working sessions, on a template pre-filled from the HR system, with scope rules fixed before the first rating. Every division was rated in 6 weeks, over the year-end break, and one data owner sent each division head their results back to confirm.

03

Calibrate twice

Then the calibration, in 2 rounds. Ratings were compared within each division, then across divisions at executive level, where any division far out of line had to explain why. Before any decision, outcomes were checked by gender, nationality and grade, so the review could not quietly favour one group. The data stayed with a closed group. Every decision could be traced to a rating, a rule and a calibration round, so it could be explained, and defended if challenged.

04

Test the rule before the names

Then the rule, tested before any name was attached. 4 ways to share the slots were weighed, from a single threshold to quotas by division size or priority, and each threshold was tested for 2 numbers: how many critical posts it kept and how many key people it lost. The recommendation was a bottom-up threshold, the highest scores extended at once, a reserve for the executive director and 6-month first extensions, so decisions stayed reversible while funding was uncertain. They reached staff through the change stream that ran alongside it.

The instrument

The post and the person

Critical post, key person. Keep both: extend, refuse the early exit, add a retention bonus if flight risk is high.

Key person, non-critical post. Keep the person, not the post: redeploy them.

Critical post, no key person. Keep the post and build the bench behind it.

Where neither is critical, the contract ends, with notice and support. Rate the post alone, and you keep the chair but lose the expertise.

What moved

What changed, and what didn't.

1 in 4key people sat in posts that a post-only rule would not have kept
Evidence Decision taken Adopted in practice Measured impact
What moved

At the recommended threshold, 1 in 4 key people sat in posts that would not have been extended. The executive director adopted the bottom-up threshold, and 4 decisions were taken on the same ratings: which contracts were extended, which applications to leave early were refused because the person was critical, which key people at high flight risk got a retention bonus, and which were redeployed rather than lost. The critical positions then became the starting point of the talent management work that followed.

What didn't move

Ratings were managers’ judgement against anchors, not validated performance or potential data, and future importance was a time horizon, not a skills forecast. Extensions had no appeal route, and whether the critical people refused an early exit stayed was beyond what we saw.

What we'd say to a peer

A voluntary separation scheme without a criticality review is a lottery: the people most able to leave are often those you can least afford to lose. Rate the post and the person separately, then test the rule on the numbers before any name is attached.

Let's talk Read more cases
Method note:

Own approach in 4 steps; 4 post types; post and person criteria scored 1 to 5 with written anchors; rater briefing pack; criticality horizon and retention risk; 2 calibration rounds; adverse-impact check; allocation options and threshold scenarios.

WeAreHuman@Work

One piece of evidence on sustainable work, every Sunday. Free, and one click to leave.

Get the weekly edition