The mechanism
The annual review has a structural flaw: the assessor and the person being assessed sit on the same side of a relationship they both want to maintain. A manager who gives a low rating then has to carry that conversation – hold a difficult development conversation, justify themselves to HR, put up with tension with someone they work with every day. It’s easier to write 4/5 and move on.
As a result, the review measures not development but the assessor’s willingness to confront – which is low almost everywhere, regardless of industry or company size.
The consequence: the company has data that looks like a development system, but doesn’t distinguish people who have genuinely changed the way they work from those who simply came across well in the appraisal conversation.
It isn’t the managers’ fault
It’s worth separating this from the search for culprits straight away. A manager who gives high ratings isn’t doing it out of ill will or a lack of assessment skills. They do it because the system forces them to: no time for an in-depth conversation, pressure to keep a good atmosphere in the team before bonus season, no tool that would give them a point of reference other than their own impression.
It’s a design flaw in the process, not a fault of the people carrying it out. And that’s why fixing it doesn’t mean “teaching managers to assess more strictly” – it means giving them data that doesn’t depend on whether they want to avoid an awkward conversation.
What does this look like in practice?
Let’s take a typical situation in a medium-sized manufacturing company – let’s call it ALUVIA. A shift supervisor gets 4.5/5 in the review cycle. On paper: the team’s star, a candidate for promotion next year. In a development interview with an external consultant, conducted independently of the line manager, something else emerges: the supervisor avoids delegating because they don’t trust the team, and solves problems themselves after hours – which for a year looks like “commitment”, but in the longer term is a guaranteed path to burnout and a bottleneck for the whole shift.
The annual review didn’t catch this, because it asked about the result (“the shift met its plan”), not about the behaviour that generated the result. The discrepancy between 4.5/5 and the real picture of behaviour only came to light when a tool independent of the manager–employee relationship came into play.
This isn’t an exception. It’s the default scenario whenever an assessment rests solely on the judgement of one person who has a reason to soften that judgement.
Three practices that fix it
1. Separate the assessment of results from the assessment of development. These are two different questions: “did they achieve the goal?” and “did they change the behaviour that leads to the goal?”. Combining them in one box guarantees that one will mask the other – a good quarterly result easily covers up a lack of change in the way someone works.
2. Add data independent of the assessor–assessed relationship. A short behavioural profile and a structured development interview conducted by someone outside that relationship don’t depend on whether the manager wants to avoid an awkward conversation. They provide an external point of reference – not to undermine the manager’s assessment, but to give them a second source they can rely on.
3. Treat the discrepancy as information, not as an error to correct. If the review says 5/5 and the behavioural data shows no change, that isn’t a contradiction to be smoothed over with a cosmetic correction of the form. It’s exactly what is worth knowing before budget and promotion decisions for 2027.
Where does NESSE come in?
These three practices sound simple on paper. In practice, companies stumble over them for one reason: they don’t have a tool that provides a “second source” without tying up months of HR’s time.
That’s why we built Development Verification – an add-on to the standard periodic review process, not a replacement for it.
It works like this:
DISC profile – a 15-minute behavioural assessment showing the assessed person’s natural working style: how they make decisions, how they react under pressure, how they communicate problems. It’s a point of reference independent of who assessed them and how.
Structured development interview (20 minutes, conducted by a NESSE consultant, not the line manager) – specific questions about decisions over the last six months, not about mood or relationships in the team. What distinguishes real behavioural change from good self-presentation shows up precisely at the level of concrete decisions, not declarations.
Comparison with the rating on the form – the NESSE consultant doesn’t re-assess, but shows where the manager’s rating and the behavioural data agree and where they diverge. The output isn’t a “corrected rating”, but a map of discrepancies for a conversation with the specific person.
For the board and HR this means one thing: instead of a distribution of stars that differentiates nothing, they get a list of people for whom it’s worth checking whether the rating reflects reality – before those people enter the promotion plan, succession plan or the 2027 development budget. That’s exactly the moment in the calendar when a mistake costs the most, because decisions made on bad data are locked in for the whole of the following year.
The most common mistake
Treating a high percentage of positive ratings as proof that development activities are effective. That reverses cause and effect – a high percentage of positive ratings more often proves the avoidance of confrontation than the quality of development.
A question for today
How many of the 4/5 ratings in your last cycle would survive a 20-minute verification conversation with someone outside the assessor–assessed relationship?
In the October issue of GROW (in Polish): the 2027 development budget – why the best budgets have fewer line items, and how to build one the board won’t cut.