The Balanced Scorecard and strategy maps

Origins and the founding claim

The Balanced Scorecard entered management practice through a 1992 Harvard Business Review article by Robert Kaplan and David Norton, and its opening premise has outlived every fashion cycle since: what you measure is what you get. Senior managers steering by financial measures alone are reading instruments that report the past — results of actions already taken — while the drivers of future performance (customer relationships, process capability, organisational learning) go unmeasured and therefore unmanaged. Kaplan and Norton’s proposal was not to replace financial measurement but to balance it: a single, deliberately compact set of measures giving a fast but comprehensive view of the business, the way a cockpit’s instruments give a pilot simultaneous readings on fuel, airspeed and altitude [1].

The balance runs along four perspectives, each framed as a question [1]:

• Financial — how do we look to shareholders (or funders)?
• Customer — how do customers see us?
• Internal business process — what must we excel at?
• Innovation and learning — can we continue to improve and create value? (Later practice standardised the label as learning and growth.)

Each perspective carries a handful of objectives and the measures, targets and initiatives attached to them. The discipline is in the word handful: the scorecard was conceived as a restraint on measurement, forcing managers to choose the few measures most critical to the strategy, precisely because organisations already drowned in reports [1].

By the mid-1990s Kaplan and Norton had repositioned the scorecard from a measurement instrument into a strategic management system: a mechanism through which organisations translate vision into concrete objectives, communicate strategy down the organisation and link it to unit and individual goals, connect strategy to budgets and planning, and — the part most often skipped — learn from measured results whether the strategy itself is working, not merely whether it is being implemented [2].

Strategy maps: the causal skeleton

The strategy map is the scorecard’s most consequential extension. Instead of four lists of measures, the map draws the organisation’s strategy as a chain of cause-and-effect relationships running upward through the perspectives: investments in people and systems (learning and growth) are hypothesised to improve specific internal processes, which improve specific customer outcomes, which produce the financial — or mission — results the organisation exists for [2].

Drawn honestly, a strategy map is a testable theory of the organisation. Drawn carelessly, it is a wish sandwich with arrows. The difference is whether each link would survive the question: what evidence do we have that improving the lower box moves the upper box, in our organisation, within our planning horizon?

The public-sector and nonprofit adaptation

The original scorecard puts financial results at the apex, because for a firm they are the point. For a ministry, county government or NGO they are not: money is an input constraint, and the apex belongs to mission. The standard adaptation therefore reorders the perspectives — mission (or the population outcome the organisation exists to change) at the top; the customer perspective recast around citizens, service users or beneficiaries; and the financial perspective demoted to an enabler alongside learning and growth.

The reordering is not cosmetic. Three design consequences follow:

1. The apex measures stop being quarterly. Mission outcomes move slowly and are co-produced with actors outside the organisation. A public-sector scorecard must therefore tolerate a slow top layer over a fast bottom layer, and resist inventing quarterly “outcome” measures that are really outputs in costume — the failure mode the results-based management tradition documents at length [4].
2. “Customer” fragments. The person who uses the service, the taxpayer who funds it and the political principal who mandates it are different constituencies with different satisfaction functions. Deciding whose perspective the second band represents is a strategic choice, not a template field.
3. Financial measures change meaning. For a firm, revenue is success; for a programme, spending is cost. Budget execution belongs in the enabling band and must never be dressed as an outcome — the same discipline the earned-value page urges from the delivery side.

Building one: objectives before measures, and few of both

The construction sequence that keeps a scorecard strategic rather than bureaucratic runs strictly in this order: objectives → measures → targets → initiatives, within each perspective, derived from an explicit strategy [2]. Organisations that start from “what data do we have” build a data-availability scorecard; organisations that start from “what must change” build a strategy scorecard and then confront their measurement gaps honestly — gaps which are themselves findings.

Hold the count down. The founding design assumed a handful of measures per perspective, and experience since has only sharpened the point: a scorecard approaching two dozen measures has stopped forcing choices, which was its entire mechanism [1]. Every measure admitted should displace a management question; a measure nobody would act on is decoration. The craft of writing the individual measures — definitions, baselines, targets, data sources — is exactly the indicator discipline covered in this site’s indicator cluster, and every test there applies to scorecard measures without modification.

The critiques, with sources

The scorecard’s dominance invited serious critique, and the strongest of it has never been refuted so much as absorbed and ignored.

The causal chain is assumed, not demonstrated. Hanne Nørreklit’s analysis in Management Accounting Research remains the standard statement: the claimed cause-and-effect relationships between the perspectives — the load-bearing feature that distinguishes a scorecard from an ad hoc measure list — are, on inspection, logical relationships asserted by the model’s authors rather than empirically validated causal ones, with time lags between cause and effect left untheorised; and the scorecard’s top-down control assumptions sit uneasily with how strategy actually forms in organisations [3]. The practical import is not that scorecards are worthless; it is that an organisation’s strategy map is a set of untested hypotheses until that organisation tests them — by examining, over time, whether its capability and process measures actually lead its customer and mission measures. Almost none do the test. An M&E-literate organisation has an unfair advantage here, since testing hypothesised results chains against evidence is the evaluation profession’s core competence.

Scorecard-as-wallpaper. A second, more sociological failure mode: the scorecard is built in a workshop, laminated, and thereafter performs the role of strategy artefact while actual decisions are made on budgets and instinct. The diagnostic is behavioural, not documentary — when did a scorecard measure last change a resource allocation? Kaplan and Norton’s own strategic-management-system framing implicitly concedes the point: the value is in the four management processes the scorecard forces, not in the document [2].

Measurement dysfunction. Any measure attached to consequences invites gaming — the target-fixation, ratchet and tunnel-vision pathologies discussed with sources on the OKR page apply to scorecard measures in full. The scorecard’s compactness is partial protection (fewer measures, watched more intelligently), but only paired counter-measures and leadership restraint about mechanical target-linking provide real defence.

Scorecard, results framework, or OKRs?

In development and public-sector settings the scorecard competes for the same institutional space as two other instruments, and choosing badly wastes a planning cycle.

Balanced Scorecard, results framework and OKRs compared

	Balanced Scorecard	Results framework	OKRs
Unit of design	The organisation	A programme or intervention	A team or unit, per cycle
Core logic	Cause-effect across four perspectives	Results chain from outputs to impact	Qualitative objective + measurable key results
Time horizon	Multi-year strategy	Programme lifetime	Quarterly to annual
Accountability weight	Internal steering	External accountability to funders	Deliberately low — stretch encouraged
Fails by	Untested causal links; wallpaper	Indicator overload; compliance ritual	Cascade fragmentation; gaming

The comparison is with, not against: an organisation can run a scorecard at corporate level while its programmes carry results frameworks and its teams set OKRs. What matters is that each layer knows which instrument governs it. The results framework itself — its structure, columns and construction — is deliberately not taught here; see the results-framework explainer at monival.com. For the OKR half of the comparison, the neighbouring page in this cluster takes over.

Checklist before adopting a Balanced Scorecard

• The organisation has an actual strategy for the scorecard to express — a scorecard cannot substitute for one.
• Perspectives are reordered for mission if the organisation is public or nonprofit, and the “customer” constituency is explicitly chosen.
• Objectives precede measures; measures stay few enough that each one is watched and acted on.
• Every strategy-map arrow is written down as a testable claim, and at least the load-bearing links have a plan for being tested against evidence over time.
• Scorecard measures meet the same definitional standards as any indicator — definitions, baselines, targets, sources.
• A named review rhythm exists at which scorecard results can change resource allocations; otherwise the wallpaper outcome is already decided.