Rule debt: 
What Happens to a Pricing Engine After 18 Months

The accumulation of rules is a cumulative process, not an isolated incident: it grows with every legitimate decision made in isolation. Five forms predominate: the exception that has become permanent, the seasonal rule that is never suspended, contradictory rules, the orphaned rule, and the redundant rule.

Before it costs you profit margin, it costs you trust: a team that no longer understands a price stops following the lead. The solution is a short cycle rather than a major overhaul, and setting an end date as soon as each rule is created.

A pricing engine that has just been deployed is a well-defined entity: rules written by people who understand why they exist, and that are consistent with one another.

Illustrative example, fictional figures: three hundred rules at the start, eight hundred eighteen months later. No one knows where the two-thirds came from anymore; some of the rules contradict each other, and a handful result in prices that no one dares to charge anymore.

The engine didn't break down—it just got clogged up. This problem has a name, a mechanism, and a solution that isn't a project but a rhythm.

Cracked, translucent layers stacked atop a jammed motor mechanism

What Is a "Rule Debt," and Why It's Invisible

A pricing rule engine operates based on a set of guidelines: minimum margin, required price differential relative to a competitor, product line consistency, rounding rules, and exceptions by supplier, region, and store format.

This park isn't a static entity. It thrives because business thrives.

Rule debt is the gap between the current set of rules as they stand and what they would be if they were rewritten today, with the benefit of hindsight. It does not consist of mistakes, but of decisions that were correct at the time they were made.

Where do these rules come from that we no longer dare to challenge?

A buyer obtains an exemption for a transaction. A regional office requests special treatment during construction work. A supplier negotiates a minimum price for its product line for a semester.

Each of these decisions is justifiable. None of them is explicitly stated to be temporary. And when reason ceases to prevail, no one is responsible for repealing the rule.

Our take: Every exemption without an end date is a hidden rule. The “rule debt” is the surplus created by the right to grant exemptions, as discussed in our article on the governance of exceptions and exemptions.

Why No One Sees Her Passing By

A rule deficit does not trigger any incidents or alerts. It manifests itself through subtle signs: a slowly rising rate of rejected recommendations and increasingly lengthy deliberations in committee.

And there’s a phrase that comes up again and again in meetings: “In any case, we’re falling behind in this category.” That’s the most serious sign.

It means that the team has recreated a parallel management system—often in a spreadsheet—that is exactly the same as the one the tool was supposed to replace.

61%

According to Cognizant, a significant portion of IT budgets is currently spent on maintaining existing systems. Executives are aiming for 27% by 2030: maintaining legacy systems remains the top priority.
(Cognizant, survey of 1,000 executives from the world’s 2,000 largest companies, reported by Le Monde Informatique, November 24, 2025)

This figure applies to information systems in general, and that is what makes it so telling: the mechanics are the same across the board. What has already been built requires more attention than what remains to be built.

The Five Forms of Liabilities

1. The Exception That Became the Norm

An exemption granted for three weeks, still in effect three years later. This is the most common type and the easiest to handle: all you have to do is make it visible.

2. The seasonal rule has never been disabled

The constraint imposed for the start of the school year, the holidays, or a weather spike—which still applies even in the off-season. We only realize it when it gets in the way—in other words, too late.

3. Rules That Contradict Each Other

Two rules that are valid on their own but incompatible for part of the catalog. The engine then makes a decision based on an order of priority that no one has consciously determined.

Illustrative example, fictional figures: One rule requires that prices remain no more than 3% above a competitor’s, while another requires a minimum margin of 25%. The purchase price rises by 6%.

The competitor doesn't budge. The two rules can no longer be satisfied at the same time: either the margin falls below 25%, or the gap exceeds 3%. The algorithm makes the choice, and no one knows what criteria it uses.

This is the most expensive approach because it produces recommendations that are difficult to understand. And a recommendation that is difficult to understand is rejected, as our article on the explainability of price recommendations shows.

4. The Orphan Rule

Its author has changed jobs, and his or her intention is not documented anywhere. No one dares to delete it, since no one knows what it protects. It remains in place as a precaution, indefinitely.

5. The redundant rule

Two rules that produce the same effect in two different ways. If you correct one, the other continues to have an effect, and you conclude that the engine is unpredictable.

The True Cost of Debt

The Cost of Trust: It Comes First

A team follows a recommendation as long as it can justify it to a buyer or store manager. As soon as it can no longer do so, it stops.

And it doesn't just stop at the faulty rule: it also considers the category and then the scope, as a precaution. Rule debt is therefore a governance issue before it is a technical one.

The cost of margin, which comes next

When two rules conflict, the engine often applies the more restrictive one—that is, the more conservative and less profitable one.

A price floor that’s been overlooked in a product line where purchasing terms have improved means profit left on the table with every sale—silently—for months on end.

The Cost of Inaction: The Heaviest Burden

A park that no one is in charge of becomes a park that no one touches. Every change requested by the business gets the same response: we don't know what it might break.

The organization then stops adjusting its pricing policy—not by choice, but because its tool has become too opaque for it to dare to do so.

79%

Some of the executives surveyed believe that their organizations will not reduce their technology debt by more than half by 2030. A liability cannot be eliminated with a single decision.
(Cognizant, same survey, November 2025)

A routine rather than a major cleanup

Natural temptation is the major overhaul. It works once, and then the system resets, because nothing has changed in the way rules are created and abolished.

Our belief: the only viable approach is a short, regular cycle aligned with the pricing cycle, with a more comprehensive review when the organization changes. When employees leave or business units merge, that’s when the rules become obsolete.

Customer Case StudyDiscount, organic, frozen foodsMulti-brand

A rhythm that prevents passivity from taking hold: each price cycle starts from a fresh baseline.

One price per store group, revised every six weeks.

The group used to manage several retail chains using Excel and manual processes. Now, prices are linked, simulated in batches, and applied by store group. Each pricing cycle starts from its own baseline.

Issue
Manage multiple franchise locations, each with its own pricing.
What Has Changed
Reference linking, batch simulations, suggested prices by store cluster.
6 weeks

by price cycle, simulated in batches across all retail chains.

Pricing Management · Multi-brand franchise group. Case study written by Booper based on an actual client project; the client has been anonymized.

The most useful criterion in this review is simple: a rule that hasn't been triggered in a year doesn't protect anything identifiable. Maintaining it as a precaution is like paying for insurance when you don't know what it covers.

The impasse is rarely technical; it is political: no one wants to be solely responsible for repealing a rule that protected something. Hence the principle:a rule is repealed by the same body that created it.

Prevent Debt from Building Up Again

The most effective measure is also the simplest: every new rule is established with an expiration date. If it is to be permanent, that fact—along with the reason for it—is stated.

The effect is mechanical: instead of justifying a cancellation, one must justify an extension. The burden of proof shifts to the other side, and that is all that matters.

Add to that a written statement of intent—who, when, why—in two lines. That’s what makes it possible, eighteen months later, to make a decision in thirty seconds instead of renewing the contract out of an abundance of caution.

Finally, a useful pricing engine must be able to show—without requiring any specific development—how many rules are active, which ones have never been triggered, and which ones conflict with one another. Otherwise, the review remains purely theoretical.

A backlog of rules is not a sign that a system is flawed. It is a sign that it is being used. The issue is not about avoiding the backlog, but about deciding how quickly to clear it.

See also in this feature

FAQ

It is the gap between the current set of rules and the one we would write today with the benefit of hindsight. It consists of decisions that were correct at the time they were made: one-time exceptions, seasonal restrictions that were never lifted, and compromises made by people who are no longer with us.

It doesn't trigger any issues: it's identified by an increase in the number of rejected recommendations and the return to parallel management in a spreadsheet.

It is a system that calculates prices based on guidelines written by the team: minimum margin, price differential to maintain relative to a competitor, product line consistency, rounding, and exceptions by supplier or region.

It is transparent and predictable, but it doesn't learn anything: its rules become outdated if no one reviews them. An AI pricing engine uses these rules as safeguards.

Based on the pricing cycle for recent rules, expired exceptions, and detected conflicts, with a broader review once a year.

A review is also necessary whenever there is an organizational change: the departure of a category manager, the merger of business units, or a new sales policy.

At a glance, this isn't realistic beyond a few dozen rules: a conflict only arises for some of the references, under specific conditions.

The solution must simulate the application of the fleet across the entire catalog, flag instances where rules conflict, and indicate which rule produced each price.

Yes, within a certain framework. A rule without a written purpose that hasn't been triggered in a year doesn't protect anything identifiable.

The decision to remove it is made in the same process as its creation, and the removal is simulated before being implemented on the shelf.

By setting an end date for every new rule, unless it is permanently in effect and justified in writing: in such cases, an extension must be justified rather than the rule’s repeal.

And by writing down the purpose of each rule in two lines: who created it, when, and why.

‍

Articles
similar
Three lines of light framing a negotiation area, with a price tag in the center
November 17, 2026
Minimum price: Give the salesperson a range rather than a specific price

The floor price is the price below which a sale cannot be finalized without approval. Together with a target price and a ceiling price, it forms a range that guides the negotiation without locking it in.

The sales rep retains their autonomy, the company maintains its margin, and sales that fall below the threshold become transparent. AI helps set these thresholds by customer segment, provided that the sales rep understands where they come from.

Read the blog post
A light strand connecting a series of price tags to decision points on a frieze
November 16, 2026
Price Traceability: Knowing Who Set This Price, When, and Why

The history shows what price was listed. Traceability shows who decided it, when, on what basis, and with what approval.

It is used to provide evidence during an audit, to justify a price to a supplier, a franchisee, or management, and, above all, to link each decision to its outcome. With AI, it distinguishes between what the machine suggests and what a human decides.

Read the blog post
Two meshing gears forced to rotate in the same direction; a spark at the point of contact
November 16, 2026
Automating a Vague Pricing Policy: AI Amplifies Your Contradictions

A vague pricing policy works as long as a human makes decisions on a case-by-case basis. When automated, it applies a set of priorities that no one has chosen, across the entire catalog and with every recalculation.

Conflicting objectives, implicit rules, and exceptions that have become the norm must be resolved beforehand. With AI, the stakes are even higher: a poorly defined objective is optimized with formidable precision.

Read the blog post
Ready to
 booster
What about your profit margins?

The pricing intelligence solution for retail leaders. Accuracy, speed, and controlled margins.

Let's discuss your pricing challenges
‍
‍