AI検索広告向けの新しいSchema.orgタグ:ブランドが押さえるべき2026年ガイド

Dev.to / 2026/5/15

💬 オピニオンDeveloper Stack & InfrastructureTools & Practical UsageIndustry & Market Moves

要点

  • Schema.orgは2026年初頭に、AdvertisedContent と SponsoredData の2つの新しい構造化データタイプを導入し、AIが有料の出稿を独立した編集的推奨として扱うことを防ぐ狙いを示した。
  • このガイドでは、ChatGPTやGoogle AI Mode、PerplexityのようなAI環境で広告を出しているブランドは、これらのタグを実装して、AIがスポンサード(広告)コンテンツをどう引用するかを制御する必要があると主張している。
  • AIモデルは、編集的なおすすめとアフィリエイトリンク、スポンサー付き商品の要素が混在したページを取り込むと、出所を信頼できる形で区別する手段を持てない場合があることを説明する。
  • AdvertisedContent は有料のプロモーション要素をHTMLブロックとして包むための仕組みであり、SponsoredData は有料または商業的な関係に由来する構造化データを示すためのものとされる。
  • 全体として、2025〜2026年に広がった「AIの回答がオーガニックの引用とプロモーションを混ぜて提示する」流れへの構造的な対策だと位置付けられている。

Schema.org rolled out two new structured data types in early 2026: AdvertisedContent and SponsoredData. These tags exist for one reason: to stop AI models from confusing your paid placements with your editorial content.

If you run ads inside ChatGPT, Google AI Mode, or Perplexity, implementing these tags is no longer optional. It is the difference between AI citing your brand as an authority and AI citing your ad as an ad.

Why Schema.org Created New Tags for AI Search

The problem started in late 2025. ChatGPT launched commerce ads. Google embedded sponsored results inside AI Mode. Perplexity experimented with promoted answers (before killing ads entirely in early 2026 to focus on subscriptions). Suddenly, AI-generated responses contained a mix of organic citations and paid promotions, and users could not tell them apart.

AI models scrape the web and ingest content. When they encounter a page that mixes editorial recommendations with affiliate links and sponsored product cards, the model has no reliable way to distinguish them. The result: AI engines started citing ad copy as if it were independent editorial endorsement.

Schema.org responded with a structural fix:

  1. AdvertisedContent: Wraps any content that is paid placement, sponsored, or promotional
  2. SponsoredData: Marks structured data that comes from a paid or commercial relationship

The Tags in Detail

AdvertisedContent

This type wraps any HTML block that is a paid promotion:

<div itemscope itemtype="https://schema.org/AdvertisedContent">
  <meta itemprop="name" content="Sponsored Product Recommendation" />
  <meta itemprop="sponsor" content="Brand Name" />
  <span itemprop="isAccessibleForFree" content="false">Sponsored</span>
  <div itemprop="item" itemscope itemtype="https://schema.org/Product">
    <span itemprop="name">Product Name</span>
  </div>
</div>

SponsoredData

This type wraps JSON-LD structured data from commercial relationships:

{
  "@context": "https://schema.org",
  "@type": "SponsoredData",
  "name": "Sponsored Product Listing",
  "sponsor": { "@type": "Organization", "name": "Brand Name" },
  "about": { "@type": "Product", "name": "Product Name" }
}

Why This Matters for GEO

GEO (generative engine optimization) relies on AI engines being able to tell the difference between genuine authority and paid promotion.

Data point: Bain research from Q1 2026 found that 80% of consumers now rely on zero-click results at least 40% of the time. Google AI Overviews drove CTR from 32% down to 16%. One brand saw a 658% traffic increase from AI citation visibility.

Implementation Steps

  1. Audit your current structured data - Run pages through Google Rich Results Test and Schema.org validator
  2. Separate editorial and sponsored HTML - Wrap sponsored sections in AdvertisedContent markup
  3. Add SponsoredData to JSON-LD - Wrap paid relationship data in SponsoredData type
  4. Validate everything - Schema.org Validator + Google Rich Results Test + manual AI citation check
  5. Monitor AI citation changes - Track whether AI engines change how they cite your content

Common Mistakes

  • Wrapping everything as sponsored - kills citation potential
  • Ignoring affiliate content - still sponsored, still needs tags
  • Using only JSON-LD - AI crawlers parsing HTML benefit from microdata too
  • One-time implementation - review quarterly as AI parsers evolve
  • Assuming AI will figure it out - pages with explicit markup had editorial sections cited 2.3x more

FAQ

What happens if I do not implement the tags? Nothing breaks immediately, but AI engines will treat sponsored and editorial content as one block. Over time, citation rates decline.

Do these tags affect Google rankings? Not directly for traditional search, but Google AI Mode uses schema signals to determine what to cite.

How long does implementation take? 4-8 hours for a standard WordPress site with 50-100 pages. 1-2 days for custom-built sites.

Should I wrap all affiliate content? Yes. Any content where you receive compensation qualifies as sponsored.

Will Perplexity benefit from these tags? Yes. Even without ads, Perplexity needs to distinguish editorial from promotional content for accurate answers.

Check your AI visibility score for free at audit.searchless.ai.

Originally published at blog.searchless.ai