68Grade C
Score 68 out of 100, grade C

Apple's homepage scores 68 out of 100 (grade C) for AI readiness. Its strongest area is crawl access (100), and the biggest drag is token economics (0). The single most valuable fix, "serve your full content in the initial html", is worth about 16 points on its own.

Hidden from AI crawlers

1%

of the text people see only appears after JavaScript runs

Reading cost

327×

~63,491 tokens of HTML for ~194 tokens of content

Structured data

3

valid schema.org entities in the raw HTML

AI crawlers blocked

0 of 11

named AI agents shut out by robots.txt for this page

Apple homepage as a person sees it, captured 3 October 2026
What people see, rendered in a real browser at 1280 px on 3 October 2026

Six pillars

Where the points go

Each pillar is scored 0 to 100 by fixed rules and weighted into the total. How scoring works

Visibility· 30%79
Actionability· 20%45
Machine facts· 20%86
Crawl access· 15%100
Token economics· 10%0
Citability· 5%55

What we found

What we would change first

  • CriticalToken economics-95 pts

    Reading this page costs an agent ~63,491 tokens for ~194 tokens of actual content, a ~327× cost multiplier (100% markup, scripts, and boilerplate).

    ~63,491 raw tokens vs ~194 content tokens, ~327× cost multiplier

  • CriticalActionability-32 pts

    136 of 257 unique rendered links have no equivalent href in the raw HTML, they only become followable after JavaScript rendering, so crawlers that skip it will miss them.

    "Kontynuuj" → https://www.apple.com/pl/ | "Mac" → /us/shop/goto/buy_mac | "iPad" → /us/shop/goto/buy_ipad | "iPhone" → /us/shop/goto/buy_iphone | "Apple Watch" → /us/shop/goto/buy_watch

  • CriticalVisibility-20 pts

    31 heading(s) shown to humans are missing from the raw HTML, they exist only after JavaScript rendering, so crawlers that skip rendering are likely to miss them.

    Quick Links | Shop Special Stores | Explore Mac | Shop Mac | More from Mac

  • CriticalActionability-20 pts

    14 distinct call(s)-to-action exist only in the JS-rendered DOM, absent from the raw HTML, so agents working from the source are likely to miss them.

    "polska", "kontynuuj", "zamknij narzędzie wyboru kraju lub regi…", "store menu", "mac menu"

Person vs AI crawler

Text people see that AI crawlers do not

1 of 42 text blocks on the rendered page are missing from the raw HTML that GPTBot, ClaudeBot and PerplexityBot receive. A few of them:

  • “Wybierz inny kraj lub region, aby zobaczyć treść właściwą dla swojej lokalizacji i zrobić zakupy przez internet.”

Fix plan

The fixes worth the most points

Ranked by how much each one alone would lift the score, as computed by the engine for this page.

  1. 01Serve your full content in the initial HTML+16
  2. 02Cut the token cost of reading this page+10
  3. 03Add the missing basic metadata (meta description)+2
  4. 04Give agents concrete facts to cite+2

How this report was made

One automated scan of https://www.apple.com/ on 3 October 2026, with the same engine (v0.3.0) as the free Siteiz scan. We fetched the page twice: as raw HTML, the way AI crawlers receive it, and rendered in a real browser, the way people see it. Scores are deterministic rules, not an AI opinion. Sites change, so this is a snapshot of that day. Methodology

How does your site compare?

Run the same scan on your own page. Free, about a minute, with your score and the fixes worth the most points.

More reports