What a MAP break actually looks like in the data

Most brands find out about a MAP break the way they find out about weather: someone mentions it. A distributor calls, a sales rep spots something on a competitor's shelf, a marketplace listing turns up in a Google search. By then the price has been live for two weeks and three other sellers have matched it.

The price was visible the whole time. What it could not tell you, on its own, is whether an agreement broke.

A price is not a violation

A listing showing €18.99 against a €24.99 floor looks like a breach. It might not be one. It could be a bundle with a different unit count, or a third-party seller you have no agreement with. It could be a promotional period you signed off on and forgot. Or the retailer's own price-match engine is reacting to someone else, which makes it a symptom rather than the cause.

To call it a violation you need four things together: the price, the seller, the moment it changed, and what it was before. Miss any one and you are guessing.

What a break looks like as an event

When a watch detects a price change, we store it as an event: the old value, the new value, the timestamp, and the seller who held the buy box at that moment. We store the transition itself, rather than a daily snapshot averaged into a chart.

That shape makes the conversation possible. "Your price on this SKU has been below the agreed floor since the 14th" is a different conversation from "we think something is off with your pricing." The first one ends with a correction. The second one ends with an email thread.

It also tells you the order of events. If three sellers dropped within an hour of each other, the first one is the one to call. The other two are almost certainly automated repricers following the leader, and calling them first wastes the week.

How fast you catch it

A break caught in an hour is a phone call. A break caught in a quarter is a renegotiation, because by then the low price has set the reference point for every buyer who saw it and every reseller who matched it.

This is why crawl frequency matters more than crawl breadth for pricing. Watching four hundred SKUs once a week gives you less than watching forty SKUs four times a day, if those forty are the ones where price discipline actually protects margin. Pick the SKUs where a break is expensive and watch those properly.

Before you build the alert

Decide what you will do when it fires. An alert nobody owns becomes an alert everyone mutes. Subscribe the person who makes the call to the seller, rather than the whole commercial team, and set the threshold at the point where you would genuinely pick up the phone. Anything smaller can pass.

Then let the rest of the price data sit in history until you need to prove what happened and when.