Six modules, one crawl, one price intelligence platform

Six modules, one collection layer. A competitor price is read once and then does six jobs: it updates a monitored listing, opens a MAP violation if it sits below your floor, feeds a repricing rule, moves your price index, adjusts a coverage and freshness number, and travels out through an integration. Nothing is collected twice, nothing disagrees with itself, and no module is sold as an add-on.

What one crawl actually changes

Most pricing stacks are assembled from parts. A monitoring tool watches competitors, a MAP vendor watches marketplaces, a repricer reads a feed, and a BI dashboard reads an export of all three. Each one collects its own data on its own schedule, so by Wednesday afternoon the MAP report and the pricing dashboard disagree about what a seller charged on Tuesday, and the meeting is spent reconciling instead of deciding.

Here the observation is the atom. One fetch produces one timestamped row holding the shelf price, was-price, stock status, shipping, promotion text, seller, currency, variant and source URL. Every module reads that same row. A MAP violation and an index point that were computed from the same observation cannot contradict each other, because there is only one number and one capture time.

The six modules of the price intelligence platform

How one observation travels through the system

This is the same competitor listing at six moments on the same morning. It is worth following once, because it explains why the modules are not really separate products.

1. 1

   Collected

   A tracked competitor page is fetched on schedule. Structured product data is read where the site publishes it, a per-domain recipe where it does not. The listing reads $159.00, in stock, free shipping, sold by DealVault.

2. 2

   Matched and validated

   The listing is joined to your SKU by GTIN first, then by title and attribute embeddings, then by image. It passes plausibility checks against its own history and against peer listings, and is written with a confidence score and a capture time.

3. 3

   Compared against MAP

   Your MAP on that SKU is $199.00, so the listing is 20.1 percent below MAP, because (199.00 - 159.00) / 199.00 = 0.201. A violation opens and the page is captured as timestamped screenshot evidence in the same pass.

4. 4

   Fed to rules

   The observation becomes an input to any repricing rule watching that competitor. Guardrails decide whether a change is even legal for you: margin floor, cost floor, MAP floor and maximum daily move are checked before a new price is proposed.

5. 5

   Rolled into analytics

   The same row moves the price index, the position mix and the category volatility figure. Because it carries a stock flag, an out-of-stock listing can be excluded from the market average rather than quietly dragging it down.

6. 6

   Sent out

   An alert goes to email or Slack, an approved price change goes to the store or a CSV, and the raw observation is available over the REST API, a webhook or SFTP. What you collect is yours to take.

Every module is on every plan

Capability is never the upsell. You pay for scale, meaning how many products you track and how often they refresh. You never pay to switch a feature on, and there is no tier where MAP evidence, repricing or the API is withheld.

The practical effect is that a small brand can run a real MAP program with screenshot evidence for $99 a month, and a retailer on Scale is buying throughput rather than permission. When you compare vendors, price the capability you need at the volume you have, not the tier name.

What the platform does not do

• It does not collect anything a shopper cannot see. No logged-in pricing, no member tiers, no negotiated B2B rates, nothing behind an account wall.
• It does not guess a price. If a page will not yield a plausible value, the observation is quarantined and the gap is shown as a gap. Charts are not interpolated to look continuous.
• It does not change prices on its own by default. Repricing proposals can go to a review queue, a CSV or the connected store. Which of those happens is your decision, per rule.
• It does not decide whether a seller is authorized. That list comes from you, because only you know who signed what.
• It does not model demand for you. There is no elasticity curve without your own sales data, and no analytics module can invent one.