Required and recommended structured data fields, how Google cross-checks rating data, the variant page decision, and the cost of copying manufacturer copy.
The product page is where both ranking and conversion are decided on an e-commerce site. This article covers title and description structure, the required and recommended fields in structured data, the scrutiny Google applies to rating data, and how to handle variant pages.
The most common mistake in product page optimisation is thinking about the page for either the search engine or the customer. In practice both want largely the same thing: clarity about what you sell, visible price and delivery information, and genuine reviews.
Title, H1 and meta description
- Page title: product name + distinguishing attribute + brand. Rather than leading with the category name, use the phrase people would actually search for.
- H1: the product name. It does not have to match the title exactly; the H1 is written for the customer, the title for the search result.
- Meta description: not a ranking factor, but it determines click-through. Put the information that drives a decision here — price range, delivery promise, return terms.
Product description: don't copy the manufacturer's text
Dozens of sites selling the same product use the manufacturer's description verbatim. That makes your page indistinguishable from hundreds of others — and gives Google no reason to prefer it.
- Who is it for, and who is it not for? "Fits small — we recommend sizing up" raises conversion and lowers returns at the same time.
- What is it used with? A cross-sell opportunity and an internal linking point.
- How does it differ from the alternatives? Keeps a comparing customer on the page.
Structured data: required and recommended fields
| Field | Status | Note |
|---|---|---|
name | Required | Product name |
image | Required | High-resolution product image |
offers | Required | Price and currency together |
brand | Recommended | Effective for products with brand recognition |
sku / gtin | Recommended | Strengthens product matching |
aggregateRating | Recommended | Enables star ratings to appear |
description | Recommended | Should be consistent with the on-page description |
Don't let the short list of required fields mislead you: the recommended fields determine how rich you look in the search result. A product whose price and availability appear directly in results earns markedly more clicks than one that does not.
Google's scrutiny of rating data
This is where shortcuts get expensive. Rating and review data must reflect real customer reviews. Google cross-checks aggregateRating values against known review sources and suppresses or penalises inflated or fabricated ratings.
- The rating should be a decimal value on a 1–5 scale (for example
"ratingValue": "4.3"). reviewCountshould be the genuine number of reviews behind the rating. A minimum number of reviews is needed for stars to appear at all.- The rating you mark up must be visible to customers on the page. A rating that exists only in structured data is a policy violation.
Because collecting reviews takes time, many stores skip this step or cut corners. The correct approach is to implement the markup on products that have reviews and omit aggregateRating entirely on those that do not. We cover the conversion impact of reviews in our conversion rate article.
Images
- The main product image above the fold is the LCP element: never lazy-load it, and load it with high priority. Detail in our page speed article.
alttext should describe the product, not be stuffed with keywords.- Use meaningful file names (
blue-leather-boot.jpg), not random character strings. - Images showing the product in use and giving a sense of scale raise conversion and bring image-search traffic.
Variants: one page or separate pages
| Situation | Approach |
|---|---|
| Variants are searched individually (e.g. "red leather jacket") | Separate indexable pages, each canonical to itself |
| Variants are not searched (e.g. size) | One page; variant selection should not change the URL, or should canonicalise to the parent |
The cost of being on the wrong side differs: unnecessary separate pages cannibalise each other, while missing pages that were needed means never capturing the demand at all.
Stock, price and delivery information
These three are both decision information for the customer and input for structured data. The value shown on the page and the value in the markup must be consistent; showing a discounted price on the page while marking up the list price loses your rich results.
Do not delete out-of-stock product pages: mark availability and show alternatives. For the detailed policy see our technical checklist.
Internal linking
A product page should not be an isolated island. From every product page, link to the parent category (via breadcrumbs), to related and complementary products, and to any blog content about that product. Those links distribute authority and keep the customer on site.
Common mistakes
- Using the manufacturer's description verbatim. It makes the page identical to hundreds of others.
- Marking up a rating that is not visible on the page. A policy violation that can cost rich results or bring a penalty.
- Lazy-loading the main image. It directly damages LCP.
- Creating a separate page for every variant. Produces dozens of near-identical pages competing with each other.
- Leaving price and stock inconsistent with the markup. Rich results are lost.
Conclusion
Product page SEO is where technical markup and sales copy meet on the same page. Writing a distinctive description, collecting genuine reviews and keeping structured data consistent with what is on the page are three jobs that raise ranking and conversion at once.
For site-wide indexing architecture, see our e-commerce SEO technical checklist.
At Commerslab we design product page templates for search and conversion together. See our e-commerce consulting service or request an audit of your product pages.
