MAP Monitoring Software: What Brands Should Evaluate

MAP monitoring tools describe themselves in similar words — monitoring, detection, enforcement, evidence — and differ in the parts the words do not cover. The questions below are the ones whose answers actually separate a tool a compliance team can stand behind from one that produces confident numbers. They are the questions we would want a prospect to ask us, and where MapProtector's answer is a limitation rather than a strength, this article says so.

How it reads Amazon

1. Through what channel does it read listing offers? The sound answer is the Selling Partner API, authorized from the brand's own Seller Central account, so the brand grants the access and can withdraw it. The other answer is scraping the storefront — and the follow-up is how it avoids being blocked. A tool that solves challenges, rotates proxies or evades detection produces observations the brand may not want to present to anyone. MapProtector reads only through the authorization the brand grants and does not run a browser against Amazon.

2. How often, and is the cadence configurable per plan? Repricers move several times a day. Ask for the real cadence, not "continuous". Ask what happens on a day a scan fails: a gap, or a compliant day?

3. What does the tool record about how much of a listing it saw? This is the question most tools cannot answer. A scan that read nine of twelve offers cannot say anything about the three it missed. If coverage is not recorded per observation, the tool will eventually report a seller as compliant because it failed to see them.

How it measures

4. Does the policy have versions with effective dates? A MAP raised in April must not turn March's observations into violations. If the tool keeps only the current floor, its history is rewritten every time the floor moves, and rewritten history is not evidence.

5. How are promotions and exceptions handled? A sale the brand approved should never reach the review queue. A distributor under negotiation should be measured but not enforced. Ask whether those are two distinct mechanisms — a promotion changes the floor for a window, an exception changes whether an observation is judged or enforced — and whether either can accidentally erase an observation. How Promotions and Exceptions Affect MAP Monitoring explains why they need to be separate.

6. What are the evaluation outcomes, and are the non-answers kept separate? Compliant and below policy are the obvious two. What does the tool do with an offer whose price it could not read, an offer a coupon made ambiguous, an offer an exception exempted, or a product no policy covered? If any of those is counted as compliant, the compliance rate is flattered.

How it treats sellers

7. What is a seller's identity? The marketplace's seller id, or the storefront name? Names change and collide; ids persist. A tool keyed on names loses history on a rename.

8. Can it mark a seller unauthorized on its own? The honest answer is no. A tool can match an observed seller against a list the brand declared and call a match authorized. It cannot know that a non-match is unauthorized; that is a decision the brand makes, with a reason, and a tool that infers it will write to the brand's own channel. Ask specifically whether "unknown" exists as a state distinct from "unauthorized". Authorized vs Unauthorized Amazon Sellers covers why this matters.

9. Does it combine authorization and compliance into a score? If so, ask how the score is computed and whether you would be comfortable quoting it to a real seller. Authorization is what the brand says; compliance is what the marketplace showed. Two facts, two columns.

What its evidence proves

10. What is an evidence package made of, and is it sealed? Ask what the record contains — seller id, exact price and currency, instant, coverage, policy version, who confirmed — and whether it is fingerprinted at sealing and verified on every read. Ask what happens when verification fails: reported as a status, or silently repaired? Ask whether the tool describes its evidence accurately — a record derived from what the source reported is not a screenshot, and a tool that calls one the other is overclaiming. What Evidence Should You Capture for a MAP Violation? sets the standard.

Who can press send

11. Can anything reach a seller without a person approving it, and from whose address? "Automated enforcement" should make a compliance team nervous. Ask whether a notice can be sent by a schedule alone, whether approve and send are separate acts, whether approval requires a fresh second factor, and what invalidates an approval given on Tuesday if the seller corrects the price on Wednesday. Ask whose address the notice leaves from: a notice from the vendor's domain misattributes the policy and misroutes the reply. MapProtector sends only from a sender the brand verified, only after a brand administrator approves and separately releases, and re-checks every blocker on the send path.

12. How is your data separated from other customers'? By application code that adds a filter to each query, or by the database refusing to return another tenant's rows? Ask whether there is any cross-customer seller reputation or score — and if there is, whose dismissals are feeding it. Ask what a vendor support engineer can see and do inside your workspace, and whether that access is time-boxed and visible.

Questions about the vendor's honesty

A short set that says a lot.

• What does the tool not do? A vendor who cannot name a limitation has not thought about the product carefully.
• Which certifications does it hold — and is the answer stated plainly rather than implied by a logo wall?
• What is the sample data on the website? Real customer data on a marketing page is a tenant-isolation question.
• Is there customer documentation you can read before buying? A tool whose behavior is only described in a sales deck is a tool you will discover in production.

MapProtector's answers are on the product page, the security page and in the documentation, including the parts that are limitations: no screenshots of Amazon, no automatic classification of sellers, no sending without a person, no certification claimed that is not held. Evaluate us against the same twelve questions.