In Brief
- OKRs do not fail because the framework is flawed — they fail because waterfall's activity-measurement defaults capture them before outcomes ever take hold.
- Evidence-Based Management (EBM) provides the outcome-measurement scaffolding that OKRs require to function — the two frameworks are complementary, not competing.
- The four Key Value Areas force a shift from "what did we deliver" to "what changed for the customer" — a shift most reporting cadences actively resist.
- Organisations that implement OKRs without changing what they reward and report are running outcome language on an activity-measurement operating system.
- The measurement problem is solvable — but the solution is structural, not a framework swap.
You implemented OKRs. The leadership team agreed on objectives. Key results were defined, dashboards were built, and the quarterly reporting cycle began. Eighteen months later, the metrics look healthier — and almost nothing has changed for customers. The teams are reporting the same data they always reported, dressed in outcome language. The executive is reading a progress update, not a measure of value created.
This is not an unusual outcome. It is the typical one. And attributing it to poor goal-setting, or to teams not understanding outcome measurement, or to OKRs being fundamentally flawed as a framework, misses the actual mechanism.
The problem is what was already in place before OKRs arrived.
Decades of waterfall project management trained every layer of most organisations — reporting cadences, tooling, management expectations, performance conversations, budget sign-off rituals — to measure activity. How much did we spend? How much did we build? Are we on track against the plan? These are not bad questions in isolation. They are the wrong questions when the objective is to understand whether investment is creating value. But they are the questions the organisation already knows how to answer, and they are the questions every system around the OKR implementation is still optimised to surface. When OKRs land on top of that infrastructure, they do not displace it. They get absorbed by it.
The result is outcome language wrapped around activity data. Teams report on features shipped, tickets closed, and roadmap progress — and describe these as key results. Objectives sound ambitious. The reporting cadence runs on schedule. And the executive reading the quarterly update has no reliable way to distinguish “we delivered value” from “we delivered output.”
Outcome language wrapped around activity data is not outcome measurement — it is the old system wearing new labels.
This is the specific failure mode Evidence-Based Management (EBM) was designed to address — not as a replacement for OKRs, but as the structural scaffolding that makes outcome measurement possible in the first place. (Scrum.org, 2024)
OKRs need scaffolding
The case against OKRs is usually made on framework grounds: the objectives are too ambitious, the key results are too vague, the alignment across teams is too loose. These are real problems, but they are symptoms. OKRs impose an obligation to measure outcomes without providing a mechanism to do so. They tell you what to aim at. They do not tell you what to measure to know whether you are getting closer.
EBM provides that mechanism through four Key Value Areas (KVAs). Each KVA describes a distinct dimension of value — one that requires a different kind of measurement to surface, and that exposes a different kind of organisational failure when it is absent.
Together, they force the shift from “what did we do” to “what changed.” That is not a framing shift. It is a structural one.
Current Value: What customers get from what exists now
Current Value (CV) measures what the organisation is actually delivering to customers today — not what it has built, but what customers use, benefit from, and would notice the absence of. The question is not “did we ship the feature?” but “are customers getting value from what exists?”
This is where most OKR implementations stall first. Because the organisation’s systems — sprint reviews, product demos, delivery metrics — are all oriented toward output, not toward whether the output is being used or valued. A team can ship twelve features in a quarter, report healthy velocity, and have zero impact on customer outcomes. CV makes that gap visible.
The game organisations play with CV is to substitute usage metrics for value metrics. Page views, login frequency, and feature adoption rates look like outcome data. They are not — they measure exposure, not value created. CV requires asking harder questions: are customers achieving what they came to achieve? What are they not doing that they need to do? Where is the gap between what exists and what would actually serve them? Those questions require qualitative investigation alongside quantitative measurement, and most reporting cadences are not built to hold them.
Unrealised Value: Value the organisation is leaving uncaptured
Unrealised Value (UV) asks what value remains uncaptured — in markets the organisation could serve but does not, needs it could meet but has not identified, and customer segments whose problems it has not yet fully understood.
This KVA does the most work at the strategic level, because it reframes the investment question. The default framing — especially under annual budget cycles — is “how do we deliver more of what we already know how to deliver?” UV demands a different question: “where is the value we are leaving on the table, and is it worth the investment and risk to capture it?” That requires market intelligence, customer discovery, and a tolerance for uncertainty that most planning cycles are designed to eliminate.
UV also surfaces the opportunity cost of technical debt and capability gaps in a way that CV cannot. A team spending a large portion of its capacity maintaining existing systems is delivering less CV than it could — and its UV is contracting, because the capacity to discover and pursue new value is being consumed by the weight of the past.
Ability to Innovate: How much the organisation can actually change
Ability to Innovate (A2I) measures the organisation’s capacity to deliver new capability. It is the internal dimension that the other KVAs depend on — because an organisation that cannot change cannot improve its Current Value or capture its Unrealised Value, regardless of how clearly those opportunities are identified.
A2I tends to be the KVA that most directly confronts the organisation’s own decisions. Every architectural shortcut, every deferred refactoring, every accumulated layer of legacy tooling reduces the speed at which the organisation can respond to what it learns. A2I makes that cost legible — not as technical debt in a developer’s backlog, but as a strategic constraint on the executive’s decision-making.
The failure mode here is measuring output velocity and calling it innovation capacity. Teams that are moving fast on known problems are not necessarily capable of moving at all on new ones. A2I requires distinguishing between throughput and adaptability — between how much the team can produce in a sprint and how quickly it can respond when the customer’s needs change.
Time to Market: How quickly the organisation learns from what it builds
Time to Market (T2M) (or Time to Value for organisations without a profit motive) measures how quickly an organisation can move from learning to delivering a response. It is the feedback loop that determines whether the other KVAs are genuinely improving or merely being reported.
Long T2M means slow learning. When it takes six months to get a change in front of customers, the organisation cannot test whether its understanding of Current Value or Unrealised Value is correct. It is operating on assumptions that are already six months stale. Shorter T2M is not a delivery preference — it is a structural requirement for any system that claims to learn from what it builds. (Forsgren, Kim & Humble, 2018)
T2M also tends to be where waterfall’s measurement defaults are most visible. Stage-gate processes, lengthy approval cycles, handoff-heavy pipelines — these were designed to reduce execution risk in predictable environments. In complex product environments, they increase the risk of building the wrong thing, slowly, at high cost. T2M surfaces that trade-off.
EBM gives OKRs the measurement infrastructure they need
The four KVAs, taken together, are not a replacement for OKRs. They are the measurement framework that OKRs require. When an organisation defines an OKR — “improve customer satisfaction in our enterprise segment” — the KVAs provide the means to answer whether the investment is actually moving toward that objective. CV tells you whether customers are getting value now. UV tells you whether you are identifying the right problems to solve next. A2I tells you whether you have the capacity to act on what you learn. T2M tells you whether you are learning fast enough to matter.
Running OKRs without this scaffolding is not running OKRs poorly — it is running OKRs on the wrong measurement infrastructure, asking outcome questions of a system equipped only to answer activity questions. The answers look like outcomes. They are not.
EBM does not solve this by replacing the framework. It solves the problem by changing what the organisation measures and, consequently, what it pays attention to. That is a more durable change than a framework swap, and a harder one — because it requires changing the questions asked in every planning meeting, every quarterly review, and every budget conversation.
What actually needs to change
Adopting EBM’s KVAs alongside OKRs will not, by itself, fix the measurement problem. The measurement problem is structural. It is embedded in what the organisation rewards and reports — in the metrics that appear on executive dashboards, in the questions that get asked in delivery reviews, in the criteria used to evaluate whether a quarter was successful.
Organisations that add outcome metrics to their existing activity-reporting infrastructure end up with more data and the same problem. The activity data is the path of least resistance. It is already there, already formatted, already understood by the people who present it and the people who receive it. Outcome data is harder to collect, harder to interpret, and harder to defend when the numbers are ambiguous — which they usually are.
The shift to genuine outcome measurement requires deciding that activity data will no longer be reported, or at least will no longer be the primary signal. That is a management decision, not a framework choice. It requires changing what gets surfaced in the reporting cadence, what questions the executive asks when the data is presented, and — most importantly — what constitutes a satisfying answer.
Organisations that make that decision find that EBM and OKRs reinforce each other cleanly. The OKRs provide direction; the KVAs provide the measurement infrastructure to know whether the direction is working. The combination produces something neither framework delivers alone: an organisational system that actually learns.
Those who do not make that decision find that the KVAs become additional reporting fields, the OKRs remain activity measures in outcome language, and the quarterly review continues to be a performance review rather than an assessment.
An organisation that continues to report activity data is not failing to implement a framework — it is succeeding at something it did not intend to maintain.
