Title & Meta Audit

Checking one page's title takes seconds. The reason metadata problems persist on most sites is not difficulty — it is that they are invisible until you look at every page at once.

A single duplicate title is harmless. Four hundred product pages sharing a templated title pattern that produces near-identical results is a structural problem affecting how the entire catalogue competes. You cannot see that pattern one page at a time, and nothing in the browser will tell you.

Auditing a crawl export surfaces the patterns. And patterns are what matter, because they usually resolve to a template rather than to individual pages.

The highest-priority finding. Two pages with identical titles are making the same claim about what they cover, which forces search engines to choose between them and often results in neither performing well. On ecommerce sites this usually comes from variant pages, paginated collections or a template that omits a distinguishing field.

Pages with no title tag get one assembled by Google from whatever it can find. Pages with no H1 give no primary statement of topic. Both are almost always template faults rather than individual oversights, which makes them cheap to fix at scale.

Lower priority than duplicate titles, since Google frequently rewrites descriptions anyway. Still worth fixing on commercially important pages, where the description is doing real work on click-through rate.

Titles that truncate and descriptions that trail off before the point. Lowest priority of the four, because a truncated title still ranks — it just converts less well than it could.

Most large sites generate metadata from templates, which is sensible — nobody is hand-writing four thousand product titles. The trap is that a template producing acceptable output for a typical product produces poor output at the edges.

Common failure modes:

• Long product names that push the template past the pixel limit, truncating the differentiating part.
• Empty fields leaving artefacts like double separators or trailing dashes.
• Variant products where the template omits the distinguishing attribute, producing identical titles across colours or sizes.
• Pagination where page two of a collection carries the same title as page one.
• Category nesting that repeats the parent category, wasting pixels on redundancy.

Auditing the output rather than reviewing the template is the only way to find these, because the template looks fine in isolation.

For templated pages, the structure matters more than the wording:

• Put the distinguishing field first. Product name before category, category before brand.
• Include the attribute that separates variants. Otherwise you are generating duplicates by design.
• Keep the brand suffix short, or drop it on deep pages where pixels are scarce.
• Handle empty fields so the template degrades cleanly rather than leaving punctuation artefacts.
• Test with your longest and shortest product names, not an average one.

Templated descriptions frequently read as filler because they are. On pages where the description matters — top collections, key landing pages, high-traffic products — write them individually. For the long tail, a clean template is acceptable, and Google will often substitute a query-matched passage anyway.

H1s belong in this audit because they interact with titles and frequently fail in the same way.

What to check: pages with no H1, pages with more than one, and pages where the H1 duplicates the title exactly. The last is not an error, but it wastes an opportunity — the title is written for the search result and the H1 for someone who has already arrived, so they can reinforce each other rather than repeat.

The most common structural fault is a theme wrapping the logo in an H1, which gives every page on the site the same H1. Worth checking before assuming individual pages are at fault.

Quarterly for a stable site. After any of these, immediately: a template change, a migration, a bulk product import, a theme update, or a new page type going live. Those four events cause the overwhelming majority of metadata regressions, and they are all easy to forget to check.

Shopify generates page titles and descriptions from theme templates and product fields, which creates a few recurring patterns worth checking.

Where a product exists as separate products per colour or size rather than as variants, the templated title frequently omits the distinguishing attribute. The result is several products with identical titles competing against each other.

Page two and beyond of a collection often inherit the same title and description as page one. Adding the page number to the template resolves it and takes one edit.

Many themes fall back to the product description's opening sentence when no meta description is set. That sentence is written for someone already on the page, not someone deciding whether to click, and it frequently starts mid-thought.

A product reachable through several collection paths can generate multiple URLs with identical metadata. Worth confirming your canonical setup handles this before treating it as a metadata problem.

The output of a metadata audit should be a short list of template changes and a shorter list of individual page edits, in that order.

1. Template fixes first. Each one resolves many pages, so the return per hour is highest here.
2. High-traffic pages next. Individually written titles and descriptions for the pages that already earn impressions.
3. Everything else last, or never. A truncated title on a page with twelve monthly impressions is not worth an editing session.

Resisting the urge to fix every row is what makes the audit useful rather than exhausting.