A strategic decision can be approved and still be reopened through everyday requests. A sales team accepts a contract outside the agreed offering. A manager borrows people from a funded initiative. An executive asks for a small addition that creates months of maintenance work. Each request may have a reasonable explanation, but together they can produce a different direction from the one the organisation agreed to pursue.
Strategy needs arrangements that make its choices hold when these pressures arise. People need to know which commitments apply, what they are authorised to decide, and how an exception can be approved. They also need confidence that the same arrangements apply when a senior leader makes the request.
Enforcement, in this sense, concerns the practical effect of an agreed decision. It includes maintaining boundaries and following through on commitments, while providing a way to challenge a decision when the evidence changes. Its usefulness depends on whether it helps the organisation carry out a sound strategy and recognise when that strategy needs revision.
Specify the Choice That Needs to Hold
Consider a software business that decides to concentrate on a standard product for regional distributors. It has been accepting bespoke reporting requests from individual customers, and maintaining those variations is consuming time needed to improve the common product. Leadership agrees that new contracts should use the supported reporting options.
The decision needs enough detail for people to apply it. Does it cover minor layout changes as well as new calculations? Can customers obtain their data for use in another reporting tool? What happens to existing contractual commitments? Who assesses a request that appears useful to the wider customer base?
Without these distinctions, the organisation can enforce different interpretations of the same statement. Sales may refuse a harmless configuration that the product already supports. Engineering may accept a new calculation because it resembles existing work, without considering its ongoing maintenance. Neither outcome follows reliably from the intended choice.
A useful boundary describes what is included, what is excluded, and which cases require a further decision. Examples help people recognise the difference. The supporting explanation should also remain accessible: the purpose is to reduce unsupported variation while improving a product that serves the chosen customer group.
That reasoning matters when an unfamiliar case arrives. Staff need enough context to apply judgement rather than treat every omission as permission or prohibition. The boundary can be specific without pretending to anticipate every possible request.
Make Resource Commitments Credible
The product decision has implications for the work already under way. Existing customers may still need support for bespoke reports, and some commitments may have to be completed or renegotiated. The common product will also need capacity if the new direction is to produce an improvement customers can see.
Leadership should therefore establish a transition plan. It might limit new variations immediately, maintain existing ones for an agreed period, and sequence the work needed to offer customers a supported alternative. The effort required to fulfil old obligations belongs in the plan alongside the investment in the future product.
A priority does not have to consume most of the budget to be credible. Maintaining a large existing service can legitimately cost more than developing a smaller new capability. The relevant question is whether the approved resources are sufficient for the commitment and whether the organisation can explain the allocation in context.
Capacity protection also needs a decision route. If a major customer incident requires people assigned to the common product, responding may be appropriate. The effect on the product commitment should then be visible, with a decision about the delivery date, scope, or replacement capacity. An interruption should not silently leave the original promise intact.
These are trade-offs that need to reach the work. A funding label or a protected budget cannot, by itself, ensure that the required people and capabilities are available when needed.
Give Exceptions a Workable Route
A standard product policy will encounter cases that deserve reconsideration. A prospective customer may request a report that addresses a common gap, or an existing obligation may require a variation during the transition. Rejecting every exception would prevent the business from using relevant evidence.
The process should establish who can approve a departure and what they need to assess. For a reporting request, that could include the customer need, expected commercial value, implementation effort, maintenance responsibility, and consequences for other commitments. The assessment should consider alternatives, including solving the need within the supported product.
Atlassian’s DACI framework separates the person driving a decision, the approver, contributors, and people who need to be informed. It offers one way to clarify participation. Those labels still need to correspond to actual authority in the business, particularly where a decision commits another team’s time.
The route must be usable before a promise is made. If an account manager cannot obtain a timely answer, the organisation creates pressure either to lose the opportunity or to commit first and seek approval later. Clear boundaries can allow routine cases to proceed locally while directing consequential departures to the appropriate owner.
Record approved exceptions and communicate them to the people affected. A one-off arrangement should not quietly become the default offering, and a pattern of similar exceptions should prompt a review of the policy. The exception process is also a source of evidence about where the current strategy may be incomplete.
Examine What the Organisation Rewards
People receive signals about priorities through more than formal announcements. Managers decide which work receives attention, what counts as a successful quarter, and whose contribution is recognised. Those signals can support an agreed direction or make it harder to follow.
For the software business, a sales target based solely on signed contract value may create tension with the standard product policy. An account manager might receive credit for revenue while the cost of a bespoke commitment appears in another department. That tension deserves examination, but it does not establish that every employee is motivated only by a metric or that the answer is a new bonus formula.
The business can clarify what constitutes an acceptable contract, include delivery feasibility in approval, and assess the quality of customer commitments during management reviews. If compensation arrangements need to change, that should be a considered change with clear expectations, rather than an assumption that one shared outcome can fairly represent every person’s contribution.
Recognition should include the work that makes the strategy sustainable. Maintaining existing customer services, reducing unnecessary variation, and identifying a costly exception can all be valuable contributions. Describing this work as unimportant simply because it is not the newest initiative would undermine the transition.
Measures also serve different purposes. The company may prioritise product adoption while continuing to monitor cash flow, service reliability, and customer obligations. Monitoring an important condition does not automatically make it an equal strategic priority. Dropping those checks to demonstrate commitment would leave leadership with less information for making sound decisions.
Respond to Departures With Evidence and Proportion
When a team acts outside an agreed boundary, the first task is to establish what happened. Was the policy accessible and understood? Did the request fall clearly within it? Had a more senior person given conflicting instructions? Was there a workable route for obtaining a decision?
Suppose an account manager promises a bespoke report after being told that the contract must close before the end of the week. The response needs to examine the manager’s actions and the pressure or instructions surrounding them. Holding only the person who signed the contract accountable would leave the wider decision process unchanged.
The immediate work is to resolve the customer commitment and prevent further misunderstanding. That may require a commercial discussion, an approved exception, or a supported alternative. The organisation should also correct the conditions that allowed an unauthorised promise to reach that stage.
A repeated, deliberate disregard of a clear and feasible responsibility warrants a different management response from an uncertain judgement made under conflicting instructions. Accountability needs that distinction. Public humiliation or automatic penalties provide little help in understanding the decision and can make people less willing to disclose problems early.
Google’s discussion of blameless postmortems illustrates, in the specific setting of service incidents, why examining contributing conditions helps an organisation learn. The same question is useful here: what information, authority, and pressures shaped the action? Answering it supports a fair assessment of responsibility and a more effective correction.
Keep the Boundary Open to Evidence
Following an agreed strategy should not require suppressing evidence against it. Staff need to be able to show that a standard report does not serve the chosen customer group, that the transition cost is higher than expected, or that a rejected opportunity reveals a valuable adjacent need.
The organisation should distinguish a request to change the decision from permission to disregard it. Until a change is approved, current commitments still guide the work. When a change is approved, the affected plans, commercial guidance, and resource allocations need to reflect it. Otherwise different parts of the business continue enforcing different versions.
A review can examine whether exceptions are rare and justified, whether the standard offering is meeting customer needs, and whether the expected maintenance benefit is appearing. Too many exceptions may indicate an unclear boundary, weak follow-through, or a mistaken assumption about the market. The evidence needs to determine which explanation fits.
This gives enforcement a practical test. Can an account manager make a reliable commitment, can delivery teams plan around it, and can an unexpected case receive a defensible decision? When those arrangements work, the strategy influences ordinary choices without requiring constant intervention. When the evidence changes, the organisation has a way to revise its course deliberately.
