The common problem
V. F. Ridgway’s foundational research on performance measurement, published in Administrative Science Quarterly, established the core mechanism behind most performance-management failure: measures, trusted indiscriminately and relied on too completely, can produce exactly the distortions they were meant to prevent — documented in his research through cases of employment interviewers and case-quota investigators, each optimising precisely for what was counted rather than for what the count was meant to represent. This mechanism is common to every organisation with a formal performance-management system. What differs sharply by context is which warning sign tends to surface first, because the specific pressures and structures different kinds of organisations operate under shape where the distortion shows up earliest and most visibly.
In a fast-growing organisation
In an organisation scaling quickly, the most common early warning sign is a widening gap between what the performance-management system was designed to measure and what the organisation now actually does — a gap that opens not because anyone changed the system deliberately, but because the work itself moved faster than the system was updated to track it. Larry Greiner’s research on organisational growth describes exactly this pattern in a broader form: mechanisms suited to an earlier stage become mismatched, sometimes quietly, as the organisation moves into a later one. In performance management specifically, this shows up as ratings, targets or review criteria that technically still apply but increasingly feel disconnected from the roles and responsibilities people actually hold — a specific and checkable symptom, distinct from more general dissatisfaction with the review process.
In a mature, stable organisation
In a more mature, slower-changing organisation, the more common early warning sign is compression: ratings, scores or assessments clustering tightly around a single middle value, year after year, with very little variance between strong and weak performers on paper. This is less a case of the system falling out of date and more a case of the system having accumulated enough administrative friction around differentiation — additional justification required for a high or low rating, for instance — that the path of least resistance for anyone using the system has quietly become uniformity. The system has not become disconnected from the work; it has become disconnected from honest differentiation within the work, which is a related but distinct failure.
In a professional or knowledge-based organisation
In organisations where the work is highly judgement-dependent and difficult to reduce to a small number of clean metrics — professional services, research, design, and similar knowledge-intensive fields — the characteristic early warning sign is a mismatch between the system’s chosen proxy measures and what actually constitutes strong performance in the role. Billable hours, output volume, or a similarly countable proxy can become the de facto definition of good performance, not because anyone believes it is the right definition, but because it is the easiest one to measure consistently. The warning sign here is subtler than compression or drift: it shows up as strong performers, by the organisation’s own informal judgement, scoring inconsistently against the formal system, and weaker performers sometimes scoring well against it — a divergence between formal and informal assessment that tends to widen quietly rather than announce itself. Where this divergence is left unexamined for long enough, it tends to produce a specific secondary symptom: the organisation’s most respected practitioners begin treating the formal system as something to be satisfied rather than something that reflects genuine judgement about their work, a distinction most of them can articulate clearly if asked directly, even though it rarely surfaces unprompted in a standard review conversation.
In a multi-site or franchised organisation
A fourth context worth treating separately is an organisation operating the same performance-management system across multiple sites or units that, in practice, differ from one another in ways the central system does not fully account for. Here the characteristic early warning sign is neither drift nor compression but divergence: sites nominally using an identical system produce results that vary sharply, and the variation is attributed, by default, to differences in site leadership quality rather than examined for a more structural cause. Sometimes the leadership explanation is correct. Often, closer inspection finds that the “identical” system is not actually experienced identically — local customer mix, staffing constraints, or operational conditions mean the same criteria land very differently from one site to the next, and a system built around a single central assumption of comparability has no way of surfacing that difference on its own.
The warning sign specific to this context is a widening gap between sites ranked well and sites ranked poorly that does not correspond to any other independent measure of site health — revenue, retention, customer satisfaction — examined separately. Where the ranking gap and the independent measures diverge, the performance system itself, rather than the sites it is ranking, becomes the more likely place to look first.
What explains the variation
Across all four contexts, the underlying vulnerability identified by Ridgway’s research is the same: any performance-measurement system, left unchecked against the reality it is meant to represent, will tend to drift. What differs is the specific pressure each type of organisation places on that vulnerability — speed of change in a fast-growing organisation, administrative friction around differentiation in a mature one, the difficulty of measuring judgement-based work directly in a professional or knowledge-based one, and a false assumption of comparability across sites in a multi-site organisation. A leadership team’s most useful first step, in any context, is identifying which of these pressures its own organisation is currently under, since the warning sign worth watching for follows directly from that answer rather than being the same checklist everywhere. None of the four is more “correct” than the others as a starting point — an organisation experiencing more than one of these pressures simultaneously, which is common in practice, may need to watch for more than one warning sign at once, rather than assuming a single diagnosis will cover everything the system is currently missing.
The same pattern appears across many kinds of organisation; see Other Sectors for how it is read elsewhere.
Questions worth considering
- Which of the performance system’s measures still match the roles and responsibilities people actually hold?
- Do ratings across the organisation vary as much as the real differences in performance would suggest?
- Where the same system is used across different parts of the organisation, does the ranking it produces agree with independent measures of how each part is doing?
More on this topic: Performance.
Your team is facing similar challenges?
Subscribe to The Riddle Ledger
Perspectives are sent as they are published, with no promotional email, and you can unsubscribe at any time.