Market Changelog / Pricing guide

Pricing guide

Monitor competitor pricing changes with evidence.

Separate a meaningful pricing transition from page noise, then give your agent a stable event and the source evidence behind it.

Updated

Pricing pages are noisy sources

A pricing page can change because of a price move, a packaging decision, a renamed plan, a temporary banner, navigation churn, or an unrelated layout edit. A useful monitoring system needs to preserve the observed transition without pretending that every difference has the same business importance.

Start with the exact public page

For the private pilot, each target is reviewed before monitoring. Good targets are public pricing, packaging, documentation, changelog, release-note, or product pages with a stable URL. Authenticated application screens and private customer data are out of scope.

Baseline firstThe first successful capture establishes the known state; it is not presented as a new market event.
Transition secondA later material difference can become a durable event with its own chg_ ID.
Evidence retainedThe source URL and bounded before/after evidence remain available for review.
Severity is a signalClassification helps prioritize review, but the agent should still inspect the evidence.

What an agent can do with a recorded price change

An agent can search a chosen time window, retrieve a high-severity pricing event, compare the stored before and after evidence, and then decide whether the change should affect a battlecard, product review, or market summary. A stable identifier lets another workflow revisit the same event later.

What the evidence does not prove

A page transition does not automatically explain why a competitor changed a price, whether every visitor sees the same offer, or how customers responded. Those are inferences that require additional sources. Market Changelog is designed to preserve what the monitored page showed and when the change was detected.

Illustrative example: a stored difference might show “Pro $49 / seat” changing to “Pro $59 / seat.” That example explains the contract; it is not a claim about a real company or customer.

Keep the target set intentionally small

The default pilot workspace is limited to five monitored targets and a bounded monthly scan allowance. That constraint makes evidence review practical while the product is still being evaluated with design partners.

Learn more about the MCP tool flow or the broader case for agent-ready competitive intelligence.

Review a pricing workflow together.

Bring up to five approved public pages and a recurring decision your agent needs to support. Onboarding is manual and the current pilot has no cost or SLA.

Request pilot access