Check product feeds against the pages customers actually see

A product description can present contradictory facts: a listing advertises one price while the selected variant has another, or an old feed describes sold stock. Before adding more copy, compare the information customers and discovery systems receive. Your merchandising and technical teams can use this review together.

Choose a sample that exposes different failure modes

Start with an available product, an unavailable item, a discounted item and a product with several variants. Include a recent catalogue change and, where relevant, a second market. Record the product and variant identifiers, destination URL, market and review time. A single successful bestseller is a weak test of a varied catalogue. The sample should explain what has and has not been checked.

Make a comparison sheet rather than a score

Use one row per sampled variant. Compare the approved source record, visible page, structured data and submitted feed. Check identity, selected option, currency, price, availability, image and destination. Add separate columns for observed mismatch, likely source, responsible person and retest evidence. Do not invent identifiers or product facts to fill a blank; ask the person who owns the catalogue.

Follow the selected variant through the journey

Open the submitted destination and confirm which size, colour or configuration it represents. Change the selection and inspect the resulting product details. Google's product-variant guidance supports describing related variants, but the implementation must reflect the actual site structure. Test an unavailable option as well as an available one. A valid markup test alone does not prove that the intended variant is correctly represented.

Repair the source and the update path

Trace discrepancies back to their origin: catalogue entry, pricing rule, template, feed mapping or update timing. Correcting an export by hand may only hide the problem until the next refresh. Assign the durable repair and agree how quickly stock or price changes should reach each surface. Review Merchant Center's actual issue details when available; do not treat every catalogue discrepancy as a confirmed cause of lost sales.

Keep discovery readiness separate from automated purchasing

Accurate public information helps people understand the offer and supports eligible shopping experiences. It does not guarantee search appearance or recommendations from every AI assistant. Product feeds and structured data also do not grant an agent payment or checkout access. Any proposed purchasing integration needs its own provider availability, authorisation, transaction testing and operational scope. This review supplies a documented content and data repair plan.

A handover checklist

Record the sampled markets, variants and review time.
Compare page, markup and feed against approved product facts.
Assign each mismatch to a source owner.
Retest after the next normal catalogue refresh.
Keep unresolved exceptions and coverage limits in the handover.

Illustrative scenario

Hypothetical example: a lamp is discounted in the catalogue, but one colour variant retains the previous price in the feed. The team records both observations, identifies an excluded variant in the export rule and corrects the mapping. After the normal refresh, the reviewer compares the same variant again. The record proves that the mismatch was repaired; it does not claim a ranking or revenue increase.

Related services and guidance

Primary references

Reviewed on 1 November 2026. The workflow and illustrative scenarios are Gavek editorial guidance; external references support the technical principles. Agree actual access, implementation, data costs and deliverables in the project scope.