All articles
Competitor monitoringEcommerceMarket research

How to build a competitor catalog monitor that survives website changes

A durable monitoring workflow for product, price, availability, and assortment changes—with evidence for every alert.

Gluecrawl · July 31, 2026 · 8 min read

At a glance

A durable monitoring workflow for product, price, availability, and assortment changes—with evidence for every alert.

  • Start with the product changes that lead to a decision, then collect only the fields that explain them.
  • Use stable identifiers, timestamps, and source URLs to separate genuine changes from collection failures.
  • Treat variants, promotions, and availability as first-class data rather than notes attached to a price.

Define what counts as a change

Competitor monitoring fails when it begins with a vague request to track everything. Begin with decisions: respond to price cuts, identify new assortment, investigate a stockout, or track promotions before a campaign. Each decision determines an alert and the evidence a reviewer needs.

A price field alone is rarely enough. Capture the regular price, sale price, currency, availability, product URL, product or variant identifier, and capture time. Add shipping threshold or promotional copy only when it changes the commercial decision.

Choose the right product identity

The matching rule is the heart of a monitor. A product URL may be enough for a small catalogue, but a SKU, GTIN, or manufacturer part number is stronger when URLs change or similar products have nearly identical names. Preserve the raw title too; it is useful when a match needs review.

Decide whether each variant is a separate row. Size, color, pack count, and subscription option can have different prices and availability. If the client buys or compares at variant level, the monitor should too.

Keep it consistent

  • New listing: a previously unseen stable identifier appears in scope.
  • Price movement: a normalized price changes beyond an agreed threshold.
  • Availability change: in stock, out of stock, backorder, or unknown changes state.
  • Coverage issue: a familiar item disappears but the page itself may have changed.

Diff normalized data, not page HTML

HTML changes constantly: tracking tags, badges, review counts, and layout adjustments can all create noise. Compare normalized fields instead. Convert currencies consistently when that is part of the brief, retain both regular and sale price, and record the raw value for auditability.

Keep the previous and current values beside the source URL. A reviewer should be able to open the relevant product page and understand why the alert exists without searching through an old spreadsheet.

Make false positives easy to investigate

Not every disappearance means a product was discontinued. It may have moved category, become temporarily unavailable, changed URL, or failed to load. Use an unknown or review-needed state before declaring a commercial conclusion.

Run a small quality sample after a site redesign or a new template appears. Check totals, duplicate products, the last page of a collection, and a handful of variants. A monitor that reports uncertainty clearly is more valuable than one that confidently reports the wrong change.

Turn alerts into a client-ready report

Group changes by priority. A weekly report might lead with meaningful price movements, new competitors in a category, and stock changes that affect a campaign; lower-priority cosmetic changes can stay in the dataset.

The deliverable becomes actionable when it includes a recommendation: match a price, update a promotion, review a new bundle, or simply watch a developing trend. The monitor supplies the evidence; the commercial team supplies the judgment.

Sources and further reading

Ready to collect the full catalogue?

Describe the fields you need and let Gluecrawl turn a website into structured data.

Get started free