Implementing and Maintaining a Pricing Policy Over Time: Governance, Exceptions, and Waivers
Fabrice Decroo
Consulting Director
August 21, 2026
Most pricing policies fail not so much because of poor design as because of a lack of active governance. Organizations that frame their automated decisions with dedicated governance are 3.4 times more effective (Gartner, 2025)—but only 28% of organizations explicitly assign this responsibility to a senior executive (McKinsey). People must steer, automation must execute—never the other way around.
Most pricing policies fail not so much because they are poorly designed as because of a lack of active governance. They are devised once, documented in a presentation, and then left to their own devices: no one is responsible for resolving unforeseen situations, and exceptions pile up without being tracked.
This guide explains how to establish a governance framework that stands the test of time—without resorting to total automation, which is not the answer.
Why Governance Is the Missing Link in Pricing Policy
The general decision-making framework for this series explains how to choose a pricing policy suited to the context. However rigorous this step may be, it is not enough: a policy is only effective if someone is responsible for enforcing it, resolving borderline cases, and revising it when the context changes.
What Fails Without Active Governance
- Exceptions are piling up without being tracked —each local manager grants their own exceptions, with no visibility into those granted elsewhere.
- A stated policy that no longer reflects the reality on the ground —the document says one thing, but day-to-day decisions say another.
- Decisions that vary from week to week on the same type of case.
- A policy that has never been revised, even though the circumstances have changed since it was last approved.
Who Manages What: Roles That Must Be Explicitly Assigned
- A policy owner —responsible for the policy's overall consistency, periodic review, and final decision-making.
- Validators by scope —authorized to approve routine exceptions, with a threshold above which they escalate the matter.
- A record of exceptions —every deviation is documented: who requested it, why, and for how long.
- A scheduled review meeting —with a set date, not left up to individual discretion.
The Real Debate: Full Automation or Human Control
×3.4 — that’s the additional efficiency factor observed in organizations that deploy a dedicated governance platform to oversee their automated decisions, compared to those that automate without such a framework, based on a sample of 360 organizations (Gartner, November 2025). Automation is not the enemy of performance; automation without governance is.
28% —that is the proportion of organizations that explicitly assign responsibility for AI governance to the CEO, and only 17% to their board of directors (McKinsey & Company, *The State of AI in 2025*). This lack of clarity at the top automatically trickles down through the organization, including into pricing.
At Booper — GENIUS Admin sets up this governance framework (permissions, rules, audit log) so that every automated decision is tracked. GENIUS Monitoring centralizes alerts so that any significant deviation is always escalated to a human before being addressed.
Exception or Waiver: The Distinction That Matters
An exception is a specific case provided for by the policy itself—a rule that explicitly allows for a deviation in an identified context. A waiver is a one-time deviation from the general rule, decided on a case-by-case basis, outside the established framework. The risk is not the exception—it is anticipated, regulated, and legitimate. The risk is the waiver that is repeated without being questioned, until it becomes an unspoken norm.
Three levels of decision-making, not just one
- Automatic — a decision made without human intervention, within a limited scope and with low stakes (low-margin transactions, minor discrepancies). Risk: silent drift if the scope is not clearly defined.
- Validated semi-automatic — automatically generated proposal, human validation before execution (significant discrepancies, sensitive categories). Risk: slowdowns if not properly equipped.
- Human decision-making — analysis and decision-making performed entirely by humans (unusual cases, high visibility, strategic importance). Risk: delays if too many cases are escalated to this level.
Proper calibration involves escalating as few cases as possible to human review, without, however, automating everything.
The pace of studying: the discipline that is most often lacking
A governance structure established once and for all is not a dynamic one. The case study of Coopérative U, the final article in this series, illustrates how this governance model plays out in a large-scale retail organization with more than 1,700 stores managed simultaneously.
A well-designed pricing policy is only valuable if it remains dynamic—constantly challenged, revised, and capable of accommodating exceptions without breaking down into silent deviations. To build this framework with our teams, discover our MPS solution : modular, managed pricing.
FAQ
This is the set of rules that define who decides what in cases not covered by the standard policy: who can approve an exception, which exceptions are tracked, and how often the policy is reviewed.
An exception is a specific case provided for in the policy itself. A waiver is a one-time deviation decided on a case-by-case basis; it must remain rare and documented, otherwise it becomes an unspoken norm.
No. According to Gartner, an organization that implements a dedicated governance framework to oversee its automated decisions is 3.4 times more likely to achieve a high level of efficiency.
A named owner, identified by type of decision. According to McKinsey, only 28% of organizations explicitly assign this responsibility to a senior executive.
By systematically documenting each exception and scheduling a periodic review. A recurring exception is a sign that the policy itself needs to be revised.
At least once a year for a formal review, and whenever there is a significant change in circumstances.
Also in this series
- Retail Pricing Policies: A Guide to Choosing the Right Strategy
- Case Study: Cooperative U—From a Reactive Pricing Policy to Predictive Management
Sources: Gartner, press release, November 4, 2025 · McKinsey & Company, *The State of AI in 2025*.
Lowering a price almost always leads to higher sales—that’s never the issue. The real question is whether the additional volume generates enough profit to offset the profit lost on each unit already sold. The answer depends on two figures that are rarely considered together: the product’s markup rate and its actual price elasticity.
In the retail sector, a product’s profitability is never fully reflected in its selling price. Part of it is determined on the shelf (the front-end margin), while another part is negotiated separately with the supplier, off the sales receipt (the back-end margin). Managing one without the other means managing an incomplete picture of profitability—and often, without realizing it, an underestimated one.
The margin, markup, and margin rate do not measure the same thing, and confusing them distorts all the resulting pricing decisions. Once these definitions and their formulas are established, the real question becomes an operational one: how can you maintain an accurate view of your margin when it changes every week, product by product, rather than recalculating it once a quarter in a spreadsheet?