Leadership announces that the company is moving upmarket. The new strategy emphasizes larger accounts, deeper adoption, and long-term customer value.
Sales is still paid for closing as many deals as possible before quarter-end. Marketing is still judged by lead volume. Product is still rewarded for shipping features requested by the largest number of users. Customer success is still measured by how many accounts each manager can carry.
Everyone can repeat the new strategy. Most people continue making decisions for the old one.
This is not necessarily resistance or poor communication. The organization has two definitions of success, and only one of them affects targets, budgets, performance reviews, and bonuses.
When strategy changes but metrics do not, the old strategy remains in charge of daily work.
Metrics are instructions disguised as observations
A metric appears to describe the business. Revenue records sales. Margin records economics. Cycle time records speed. In practice, the selected metric also tells people what deserves attention.
Attach a target and a reward, and the message becomes stronger. The organization learns which tradeoffs are acceptable, what evidence wins an argument, and which problems can be ignored until later.
This is why measurement systems are part of strategy rather than a reporting layer beneath it. Research on management control systems describes a two-way relationship between controls and strategy: strategy shapes the controls, but the controls also influence which strategy emerges and survives.
Consider a software company shifting from self-serve customers to complex enterprise accounts. The old acquisition model may reward rapid signup growth. The new model depends on security approval, implementation quality, multi-team adoption, and renewal confidence. If the executive meeting still opens with monthly signups, those slower capabilities remain secondary, whatever the strategy deck says.
The number does not have to be wrong to be governing the wrong business.
The old strategy survives in ordinary decisions
Strategic conflict rarely announces itself as conflict. It appears as reasonable local choices.
Marketing buys cheaper leads because cost per lead is on the dashboard. Sales accepts a poor-fit account because the contract helps the quarter. Product prioritizes a visible feature that supports demos over an unglamorous permission system required for deployment. Operations protects utilization by avoiding the spare capacity needed to learn a new service model.
Each team can defend its decision with approved evidence. Together, those decisions reproduce the previous strategy.
The persistence is strongest when metrics are connected to consequences. A leader can ask for collaboration, customer quality, or experimentation. If promotion depends on individual output, people receive the real message. A company can say that recurring revenue matters more than project volume. If budgets are allocated from last year’s billings, the legacy offer keeps its political advantage.
This is not a problem solved by adding one new KPI to an already crowded dashboard. New measures often arrive without old measures losing authority. The company then asks teams to optimize growth and margin, speed and learning, standardization and customization, short-term output and long-term capability, all at once.
The result is not balance. It is ambiguity settled by whichever metric has the clearest reward attached to it.
Successful history makes the scoreboard harder to change
Old metrics often earned their position. They helped the company grow, controlled risk, and made performance legible. That history gives them credibility long after the environment changes.
The danger is not merely that teams game a target. It is that the measurement system narrows what management can see. A metric built for yesterday’s advantage can continue reporting good news while the capability underneath it deteriorates.
Research into the decline of Nokia Mobile Phones describes this mechanism in detail. Growth, scale, cost efficiency, demanding targets, and established indicators reinforced the prevailing managerial view, while important deterioration was not captured. The study argues that management controls can reinforce cognition and contribute to organizational inertia, even when they were designed to support performance.
The lesson is not that Nokia failed because of one KPI. It is that a coherent measurement system can become self-confirming. Leaders see the business through measures chosen by the existing strategy, then use those measures as proof that the strategy still works.
Changing the words at the top does not immediately change that loop.
A strategic transition needs two forms of truth
Replacing every old metric would be reckless. The existing business still pays salaries, serves customers, and funds the transition. Financial controls do not become irrelevant because the company has a new ambition.
The challenge is to distinguish constraints from definitions of success.
Cash, safety, compliance, and service commitments may remain non-negotiable constraints. A volume target, utilization assumption, or lead definition may belong specifically to the old model. Keeping the former protects the company. Keeping the latter can prevent the new strategy from becoming real.
During transition, management needs two forms of truth. One shows whether the current engine remains healthy enough to fund the change. The other shows whether the new engine is becoming credible.
The second set may initially look less like polished outcomes and more like evidence of capability: time to deploy, adoption across roles, repeatable implementation steps, willingness to pay at the new price, renewal quality, or the proportion of work that can be delivered without exceptional intervention.
These measures are not substitutes for revenue. They make visible the conditions from which future revenue is supposed to come.
Change the conversation attached to the number
A metric migration starts by naming the choices the new strategy requires. Which customers should now receive priority? What work should the company stop accepting? Which capability must improve before growth is safe? Where is the business willing to perform worse temporarily in order to build a different advantage?
Then inspect the operating system around those choices: dashboards, targets, incentive plans, planning templates, investment gates, meeting agendas, and promotion criteria. A metric can disappear from a strategy document while remaining powerful in all seven places.
Robert Simons’s longitudinal work on strategic renewal found that control systems can do more than monitor results. Leaders used them to formalize beliefs, set boundaries, define critical performance variables, and focus debate on strategic uncertainty. The implication is practical: measures should not merely confirm execution. They should help the organization learn whether the strategic assumptions are becoming true.
That requires review, not just replacement. Some measures should be retired. Some should remain as guardrails. Some should be segmented so that performance in the old and new models is no longer averaged together. Some should be explicitly temporary, because their job is to guide a transition rather than define success forever.
Most importantly, conflicts should be visible. If the new strategy requires fewer, better-fit customers while compensation rewards deal count, leadership should not ask teams to “balance” the contradiction. It should decide which behavior the company is prepared to fund.
A strategy becomes operational when a capable person can follow it and still be considered successful. Until the measurement and reward system makes that possible, the strategy is competing with the company that already exists.
The old strategy does not need defenders. It has dashboards, routines, budgets, and bonuses. A new strategy takes charge only when the organization changes what counts as winning.
