Ganesh HS ·
At the year-end review, every department head presented a scorecard with most indicators green. Sales had hit calls made and proposals sent. Operations had hit orders processed and tickets closed. HR had hit positions filled and training hours delivered. The managing director then presented the year: revenue below plan, two large clients lost, margin down. Nobody in the room could see how a business in which every team had succeeded had failed, and that inability was the finding.
The scorecards were not wrong. They were measuring what each team did rather than what the business needed from it, and a team can do a great deal of the former while producing none of the latter.
Almost every KPI framework in a growing business is built from activity measures, because activity is easy to count and comfortable to report. Calls made, hours delivered, tickets closed. Activity measures answer 'were people busy'. The business question is 'did it work', and that requires outcome measures — revenue retained, orders delivered on time and complete, people who stayed past a year — which are harder to attribute and less flattering, which is why they are avoided.
ACTIVITY proposals sent easy, comfortable, meaningless
OUTPUT proposals accepted better, still not the point
OUTCOME revenue from new clients what the business needed
A framework built on the top row lets every team
succeed while the business fails. Build from the
bottom row and work upward only where you must.Working upward matters. Some roles are far enough from the outcome that they need an output measure with a stated link to it — a recruiter cannot own revenue, but can own time-to-productive-hire, which the business has agreed feeds it. The discipline is that every measure on every scorecard can be traced, in one or two steps, to one of the handful of numbers the business actually runs on. If it cannot, it is decoration.
The usual build is bottom-up: each manager proposes indicators for their team, HR collates them into a framework, and the result is a hundred measures with no relationship to each other. Reverse it. The leadership names the three or four outcomes the year depends on — usually some form of revenue, margin, retention and delivery — and each function is asked which of those it moves and how. The scorecard for a team is the answer to that question and nothing else.
Three or four. If the leadership cannot agree what they are, no framework will fix that, and the exercise has found the real problem.
Which outcome does this team move, and through what mechanism. Some teams move one directly; most move one through an output that has to be named.
The KRA is the responsibility in words — 'retain the clients we have'. The KPI is how we would know — 'revenue retained from last year's clients'. Frameworks that list KPIs without KRAs produce numbers nobody can explain.
Three to five measures per role. A person with twelve indicators optimises none of them, and a manager reviewing twelve reviews none of them either.
A scorecard reviewed once a year is a scorecard nobody manages against. The number is discovered at appraisal, explained, and forgotten. The same measures reviewed monthly — fifteen minutes, the number and the one thing being done about it — become the way the team is actually run. This is where most performance and productivity consulting work lands: the framework is a page, and the cadence is the change.
Monthly review also exposes bad measures quickly. An indicator nobody can move in a month is either an outcome that belongs a level up, or a lagging measure that needs a leading one beside it. An indicator that is green every month regardless of effort is measuring something that does not vary, and should go.
Most outcome measures are lagging: revenue retained is known after the client has already left, attrition after the person has gone. A scorecard made only of lagging measures tells a team it failed three months after the decisions that failed it, which is too late to be useful. Each lagging outcome needs a leading indicator beside it — something observable now that has been shown, in this business, to move the outcome later. Client contact frequency before retention; exit-interview signals before attrition; on-time inputs before a clean payroll month.
The leading measure is the one the team can act on this month, and the lagging measure is the one that proves the leading measure was the right choice. Where the two stop moving together, the leading indicator was a guess and needs replacing. That test — does the leading number predict the lagging one — is worth running every quarter, and it is what keeps a framework honest rather than merely tidy.
Choosing leading indicators is the part that needs judgement rather than a template, because the right one depends on how this particular business loses clients or people. It is usually where an outside eye earns its fee: not in designing the scorecard, but in finding the two or three early signals that this business has never noticed it could watch.
Any measure that determines pay or standing will be optimised, and people are ingenious. Calls made rises while conversations shorten. Tickets closed rises while reopened tickets rise faster. The defence is not moral exhortation; it is pairing. Every volume measure gets a quality measure beside it that the same behaviour would damage — closed tickets paired with reopen rate, proposals sent paired with acceptance rate — so that gaming one shows up in the other.
A KPI framework sits between two other things and fails when either is missing. Above it, the business plan supplies the outcomes; where there is no plan, the framework has nothing to trace to, and HR strategy and workforce planning is the prior conversation. Below it, the appraisal uses the measures; where the performance management process is a form completed annually, the framework is reviewed annually too, and the monthly cadence that makes it work never happens.
Built top-down, capped at a handful per role, paired against gaming and reviewed monthly, a framework does the one thing the opening room needed: it makes it impossible for every team to succeed while the business fails, because the teams' measures are the business's measures, one step removed. That alignment is what performance and productivity consulting exists to build, and it is almost always shorter than what it replaces.
A KRA is a responsibility stated in words — what the role exists to achieve. A KPI is the indicator by which you would know it was achieved. Every KPI should hang off a KRA; a KPI with no KRA is a number nobody can explain.
Three to five. Beyond that nothing is prioritised, and the monthly review that makes the framework work becomes impossible to hold in fifteen minutes.
They can be, once they have run for two or three cycles and been shown to measure what matters. Linking pay to a new, untested measure is the fastest way to discover it can be gamed.
Give them an output measure with a stated link to an outcome — time to productive hire, first-pass accuracy — and write the link down. The trace is the point, not the distance.
The measures monthly, in the team. The framework itself annually, retiring measures before adding any.
Structure, roles, policy, performance and reward.
Speak with an HR Management ConsultantMost frameworks are forty pages nobody opens. What a usable one contains, how to build it from the roles you already have, and how it connects to hiring, appraisal and progression.
Read the articleThe values statement is on the wall and the shortlist has not changed in three years. Where hiring, development and complaint handling quietly narrow, and how to make those decisions consistent without a programme.
Read the articleA flat structure works until the founder becomes the only route to a decision. How to tell when you have passed that point, and how to redesign without a reorganisation.
Read the articleIdentify critical compliance gaps, payroll leaks, and hiring bottlenecks with a free 360-degree HR audit. Get a clear roadmap for your business.
Prefer the full picture? Request a free 360-degree HR audit.