Structured data (JSON-LD)

Schema markup: /resources/guides/seo-audit

Page fetched 11 September 2026 (HTTP 200, 48,572 bytes, canonical self-referencing). Everything below is read from the served HTML, not assumed.

Initial assessment

Page type. Long-form guide. Visible H1 How to run an ecommerce SEO audit; <title> differs (Ecommerce SEO Audit: The One Part With a Published Pass Mark | Addition). Byline line reads By the Addition team · Updated 7 September 2026 · 12 min read. No named person is credited, and no first-published date is shown anywhere on the page. Five <details> Q/As sit under the H2 Questions people ask (#faq). Eight <img> elements in the body, all with alt text except the small icon; two of them are 1720x986 and 2519x1623 (the rest carry no width/height attributes). <html lang="en">.

Current state. The page already ships one application/ld+json block, an @graph of three nodes:

Homepage JSON-LD already declares Organization at https://addition-labs.com/#organization (logo https://addition-labs.com/logo/addition-mark-white.png, resolves 200) and WebSite at https://addition-labs.com/#website. The guide should reference those @ids rather than redeclare a loose Organization.

Goal. Article rich result on desktop and mobile, breadcrumb on desktop, and no markup that claims a rich result Google has stopped serving.

Rich result eligibility (checked against Google's own documentation on 11 September 2026)

Author markup: the page credits "the Addition team", so author stays an Organization. Do not invent a Person. If a named author is added to the page later, switch to Person with a profile URL.

Schema implementation

Replace the existing <script type="application/ld+json"> with this block. Every value below is taken from the page as served; the two items marked TODO need a decision from you.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Organization",
      "@id": "https://addition-labs.com/#organization",
      "name": "Addition",
      "url": "https://addition-labs.com/",
      "logo": {
        "@type": "ImageObject",
        "url": "https://addition-labs.com/logo/addition-mark-white.png"
      }
    },
    {
      "@type": "WebSite",
      "@id": "https://addition-labs.com/#website",
      "name": "Addition",
      "url": "https://addition-labs.com/",
      "publisher": { "@id": "https://addition-labs.com/#organization" },
      "inLanguage": "en"
    },
    {
      "@type": "WebPage",
      "@id": "https://addition-labs.com/resources/guides/seo-audit",
      "url": "https://addition-labs.com/resources/guides/seo-audit",
      "name": "Ecommerce SEO Audit: The One Part With a Published Pass Mark | Addition",
      "description": "Most audit checklists mix one thing Google publishes a threshold for with fifty things it does not. Here is the difference, and the ecommerce-only section that generic audits leave out entirely.",
      "isPartOf": { "@id": "https://addition-labs.com/#website" },
      "inLanguage": "en",
      "primaryImageOfPage": { "@id": "https://addition-labs.com/resources/guides/seo-audit#primaryimage" },
      "breadcrumb": { "@id": "https://addition-labs.com/resources/guides/seo-audit#breadcrumb" }
    },
    {
      "@type": "ImageObject",
      "@id": "https://addition-labs.com/resources/guides/seo-audit#primaryimage",
      "url": "https://addition-labs.com/img/proof/psi-failed-assessment.png",
      "contentUrl": "https://addition-labs.com/img/proof/psi-failed-assessment.png",
      "width": 1720,
      "height": 986,
      "caption": "A PageSpeed Insights report showing a failed Core Web Vitals assessment, with LCP red and INP and CLS green"
    },
    {
      "@type": "Article",
      "@id": "https://addition-labs.com/resources/guides/seo-audit#article",
      "headline": "How to run an ecommerce SEO audit",
      "description": "Most audit checklists mix one thing Google publishes a threshold for with fifty things it does not. Here is the difference, and the ecommerce-only section that generic audits leave out entirely.",
      "image": [
        "https://addition-labs.com/img/proof/psi-failed-assessment.png",
        "https://addition-labs.com/img/proof/gsc-catalogue-flat-clicks.png"
      ],
      "datePublished": "2026-07-06",
      "dateModified": "2026-09-07",
      "author": {
        "@type": "Organization",
        "@id": "https://addition-labs.com/#organization",
        "name": "Addition",
        "url": "https://addition-labs.com/"
      },
      "publisher": { "@id": "https://addition-labs.com/#organization" },
      "mainEntityOfPage": { "@id": "https://addition-labs.com/resources/guides/seo-audit" },
      "isPartOf": { "@id": "https://addition-labs.com/resources/guides/seo-audit" },
      "inLanguage": "en",
      "articleSection": "Guides"
    },
    {
      "@type": "BreadcrumbList",
      "@id": "https://addition-labs.com/resources/guides/seo-audit#breadcrumb",
      "itemListElement": [
        { "@type": "ListItem", "position": 1, "name": "Home", "item": "https://addition-labs.com/" },
        { "@type": "ListItem", "position": 2, "name": "Resources", "item": "https://addition-labs.com/resources" },
        { "@type": "ListItem", "position": 3, "name": "Guides", "item": "https://addition-labs.com/resources/guides" },
        { "@type": "ListItem", "position": 4, "name": "How to run an ecommerce SEO audit" }
      ]
    },
    {
      "@type": "FAQPage",
      "@id": "https://addition-labs.com/resources/guides/seo-audit#faq",
      "mainEntity": [
        {
          "@type": "Question",
          "name": "Can I do my own SEO audit?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "Yes for the parts with published thresholds, and it is worth doing. Core Web Vitals is the one group with a pass mark somebody else set. Two more checks need no judgement either, because each answer is a count you compare against your own intention. Core Web Vitals in field data. An indexed-pages count against what you meant to index. And a read of your robots.txt. The judgement layer is harder. Which filter combinations deserve to exist. Which category pages compete with each other. And which of the hundred remaining checklist items matter for your particular catalogue. That is experience, not tooling, and no crawler outputs it."
          }
        },
        {
          "@type": "Question",
          "name": "Can ChatGPT do an SEO audit?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "It can read a page you paste in and tell you sensible things about it. It cannot crawl your site, see your field performance data, or know which category pages cannibalise each other. All three need access to data it does not have. So it is useful for the per-page layer and not for the site layer. Anyone selling you an AI-generated site audit is selling you a checklist with your domain name typed into it."
          }
        },
        {
          "@type": "Question",
          "name": "How much does an SEO audit cost?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "No published survey prices audits specifically. The nearest published figure is Ahrefs' survey of 439 providers, which puts ongoing SEO at a $2,917 average monthly retainer and $111 an hour. What can be said is what moves the number. Catalogue size dominates. Whether implementation is included roughly separates a document from a fixed site. And whether anything is measured afterwards decides if you can ever tell it worked. Ask those three questions before asking the price and the quotes become comparable."
          }
        },
        {
          "@type": "Question",
          "name": "What are SEO audits?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "A structured check of whether search engines can crawl your site, index the pages you want indexed, and rank them for the terms that matter. For an ecommerce store the specific weight falls on product pages, category pages and the filter combinations they generate. The useful ones sort their findings into three. Those with a published threshold, which is essentially Core Web Vitals. Those where Google documents the mechanism but never states a number, which is where crawling, canonicals and faceted navigation sit; and the ones that are convention. The order of work follows from that separation."
          }
        },
        {
          "@type": "Question",
          "name": "What does faceted navigation mean?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "The filter sidebar on a category page: size, colour, price, brand. Each selection typically produces a new URL, and combinations multiply, so three groups of three options can generate 27 addresses from a single category. Google documents the problem this creates. Crawlers cannot tell a useful new URL from a useless one without fetching it, so time spent on filter combinations is time not spent discovering your new products."
          }
        }
      ]
    }
  ]
}
</script>

What changed versus the block on the page today

1. Article.image added: two crawlable in-body screenshots, both wider than 1200px, both with descriptive alt text on the page. The site-wide og-image.png (1200x630) was deliberately not used because I did not inspect what it depicts and Google asks for an image that represents the article rather than a logo.
2. Organization and WebSite nodes carry the same @ids the homepage already uses, so publisher, author and logo resolve on this page without a second, conflicting Organization.
3. WebPage and ImageObject nodes added so the Article has an explicit isPartOf, primaryImageOfPage and breadcrumb relationship. Optional, but it is what the Rich Results Test and most graph-aware consumers read cleanly.
4. Last breadcrumb item no longer repeats its own URL (Google's example omits item on the current page; either form validates).
5. articleSection: "Guides" mirrors the visible Guide label above the H1.
6. FAQ answer text kept verbatim, paragraph breaks collapsed to a space. If you prefer, Answer.text may carry <p> tags; both validate.

TODO: two values you need to confirm

• datePublished: 2026-07-06 is carried over from the existing markup. The page shows only Updated 7 September 2026. If 6 July 2026 is the real first-publish date in your CMS, keep it. If it is not, remove datePublished rather than guess; a wrong date is worse than none. Google also recommends a time and timezone (2026-07-06T09:00:00+00:00) if you have one.
• Visible breadcrumb. Open the page in a browser and check the JS-rendered header. If it shows Home > Resources > Guides, the BreadcrumbList node is fine. If it does not, either add a visible trail (a small <nav aria-label="Breadcrumb"> above the H1 is enough) or delete the node and the WebPage.breadcrumb reference.

Implementation note

The served HTML is a static build (hashed /assets/*.js bundles, JSON-LD inlined server-side in <head>), so the block above drops in as a template partial for the guide layout; no client-side rendering is needed, and the FAQ answers should be pulled from the same source that renders the <details> elements so the two cannot drift.

Testing checklist

• [ ] Validates in Rich Results Test (https://search.google.com/test/rich-results): expect Article and Breadcrumbs detected; FAQ will be listed as a detected type but is not a served rich result.
• [ ] Schema.org validator (https://validator.schema.org/) reports no errors.
• [ ] Matches page content: headline = H1, dateModified = Updated 7 September 2026, five FAQ pairs = five <details>.
• [ ] Both image URLs return 200 and are not blocked in robots.txt (checked: 200, robots.txt is Allow: /).
• [ ] datePublished confirmed or removed (see TODO).
• [ ] Breadcrumb trail visible in the rendered header (see TODO).
• [ ] After deploy: Search Console > Enhancements shows Article and Breadcrumbs with no errors; do not expect an FAQ enhancement report.