Partly, and the parts they get wrong are the parts that matter most. We read the machine-readable markup on 108 fetches (107 distinct pages) across faymas.in, faymas.net, faymas-in.com and faymas.com.in on 19 October 2026. 105 of them declare exactly one JSON-LD block and 3 declare none. Twelve of those pages declare an FAQPage carrying 65 questions — all 65 question strings appear as text a reader can see, and 0 of the 65 answers do, because every answer lives inside script data that never renders. faymas.com.in marks its privacy policy, about page and contact page as Article. And not one of the 44 faymas.net pages we measured carries a published or modified date.
This guide analyses published prompt text. No generation tests were run.
What we measured
One script, one measure(), four domains, 19 October 2026. For each page it fetched the server HTML, pulled every <script type="application/ld+json"> block, parsed it, and walked the whole object tree collecting every @type it found. Then it stripped script and style blocks, removed tags, and searched the remaining readable text for the strings the markup had just claimed.
The sample was not drawn the same way on each domain, because the domains are not shaped the same way. faymas.in publishes a 19,550-URL sitemap; we took its homepage, 10 of its 26 blog posts, 30 random prompt pages, 6 collection pages, 4 category pages and 4 static pages = 55. faymas.net declares 41 URLs; we fetched 19 of them (18 distinct — the apex and the www home were counted separately), plus 25 randomly chosen pages from the 213 /prompts/p/ URLs its own internal links point at but its sitemap omits. faymas-in.com has 3 posts and faymas.com.in has 4 pages; we took all of them.
Who declares what
| Domain | Pages fetched | With 1 JSON-LD block | With none | Types seen most often |
|---|---|---|---|---|
| faymas.in | 55 | 54 | 1 (/copyright) | BreadcrumbList 53, ImageObject 41, Person 40, CreativeWork 30 |
| faymas.net (declared) | 19 | 17 | 2 (/about, /contact) | FAQPage 8, SoftwareApplication 8, CollectionPage 6, CreativeWork 6 |
| faymas.net (unlisted gallery) | 25 | 25 | 0 | CreativeWork 25, BreadcrumbList 25, Person 25 |
| faymas-in.com | 4 | 4 | 0 | WebSite 4, Person 4, BlogPosting 3 |
| faymas.com.in | 5 | 5 | 0 | Article 5, Organization 5, WebSite 5 |
Two things stand out before any interpretation. First, the markup is present almost everywhere: 105 of 108 fetches carry a block, and not one carries more than one. Second, the three pages with nothing are the pages that tell you who is behind the site — faymas.in's copyright page and faymas.net's about and contact pages. The pages that sell have markup. The pages that explain do not.
The FAQ answers that are not on the page
This is the sharpest finding, and it runs in the same direction on both large domains. Twelve pages declare FAQPage: four faymas.in collection pages and eight faymas.net model pages. Between them they carry 65 questions with 65 answers.
We tested each string three ways. In the full HTML, 65 of 65 answers appear. Strip out only the JSON-LD block and search again: still 65 of 65, so the answers are not merely a copy of the markup. Strip out every script and style block, remove all tags, and search the text a reader would actually see: 0 of 65.
The questions behave the opposite way. All 65 question strings survive the same stripping. So the pattern is precise: the question is rendered, the answer is not. Whatever component renders an FAQ accordion on these two sites ships the question into the DOM and leaves the answer in the payload.
That matters because an FAQPage block is a promise about the page. It says "this page answers these questions." If the answer text is only reachable after JavaScript runs, the promise is kept for a browser and broken for anything that does not execute scripts. We did not test what Google specifically renders here — we measured server HTML only, and we did not check whether the answers appear in a rendered DOM.
Dates: two opposite failures
| Domain | Pages with datePublished | Values | Same day as dateModified |
|---|---|---|---|
| faymas.in | 40 of 55 | 2026-07-17 … 2026-10-17, spread over 8+ distinct days | 10 of 40 |
| faymas.net (declared) | 0 of 19 | — | — |
| faymas.net (unlisted gallery) | 0 of 25 | — | — |
| faymas-in.com | 3 of 4 | 2026-09-23 (×2), 2026-10-03 (×1) | 1 of 3 |
| faymas.com.in | 5 of 5 | 2026-09-30 on all five | 3 of 5 |
faymas.net declares no date anywhere. Forty-four fetches, zero published dates, zero modified dates. Its 25 unlisted gallery pages each declare a CreativeWork with an author and a breadcrumb, and no indication of when the thing was made.
faymas.com.in declares a date everywhere, and that is worse. All five pages claim 2026-09-30 — the homepage, the about page, the contact page, the privacy policy and the duplicate homepage fetch. A privacy policy that was "published" the same day as the homepage is not a publication date. It is a build timestamp, and a build timestamp is what a sitemap generator writes when nothing else is available.
faymas.in is the only one of the four where the dates look like dates. They spread from mid-July to mid-October 2026 and they differ page by page.
Declared type versus the page you actually get
Three mismatches, none of them subtle.
faymas.com.in calls its boilerplate articles. Its about page, privacy policy and contact page all declare @type: Article with a headline, an author, a publisher and an image. Three of the five pages we measured on that domain are pages no editor would call an article. This is not a judgement call about a borderline page — a contact form with six input controls is not an article.
faymas.net calls its forms software. Eighteen of the 19 declared fetches carry at least one input, textarea or select control. Eight of them declare SoftwareApplication. faymas.net does not build Gemini, GPT Image, Grok Imagine, Kling or HappyHorse; it wraps them. Declaring the wrapper as the application is defensible in schema terms for a tool page, and it is also the cheapest possible reading of "what is this page."
faymas.in's collection pages undercount their own contents. Five /prompts/ pages declare an ItemList with numberOfItems: 24. The same pages link to 37, 37, 38, 38 and 37 distinct /prompt/ URLs. Its four category pages, by contrast, declare 20 and link to exactly 20. The category pages count correctly; the keyword pages do not.
What the markup drops that the title keeps
On faymas.in, 40 of 40 pages that carry a schema name have a <title> that is that exact name plus a | … suffix — usually | Faymas Blog. The brand signature lives in the title tag and is stripped out of the structured data. On faymas.net the two agree more often: 8 of 14 pages with a schema name match their title exactly. On faymas.com.in, 3 of 5.
Read together with the faymas.in homepage block, the picture is consistent. That homepage declares a WebApplication named Faymas, category ProductivityApplication, operating system All, with an Offer at price 0 USD. The machine-facing self-description is a free productivity app. The brand word that appears in every title tag appears in the markup only as that app's name.
What this check cannot tell you
Server HTML only. Every page here was read with a plain HTTP request; nothing was rendered in a browser. If an accordion opens on click and injects its answer text at that moment, we would not see it, and the 0-of-65 answer figure would describe the pre-render state rather than the page as experienced.
We read markup, not effect. Declaring FAQPage with unrendered answers does not tell you whether a search engine shows rich results for these pages. We did not query a search engine, and we did not measure impressions, rankings or traffic. No such data appears anywhere in this article.
Sample, not census. Thirty of 19,490 faymas.in prompt pages and 25 of 213 faymas.net gallery pages were measured. The two small WordPress domains were measured in full because they have almost nothing to measure: 3 posts and 4 pages between them.
Snapshots move. A previous count of faymas.net's unlisted gallery pages on 17 October 2026 found 315 distinct URLs; this run found 213 from the same method. The link sets are not stable, so treat the gallery figures as a sample of a moving target.
FAQ
Do the Faymas sites use structured data at all?
Yes, nearly everywhere: 105 of 108 fetches across the four domains carry exactly one JSON-LD block, and none carries more than one.
Which Faymas pages publish no structured data?
Three of the 108: faymas.in's copyright page and faymas.net's about and contact pages.
Are the FAQ answers on these pages readable?
No. Across the 12 pages declaring FAQPage, all 65 question strings appear in readable text and 0 of the 65 answers do; the answers exist only inside script data.
Which faymas domain declares no publication dates?
faymas.net. Zero of the 44 pages we measured, declared and unlisted alike, carry a datePublished or dateModified value.
Does faymas.in declare the right number of items on its collection pages?
Not on the keyword pages. Five /prompts/ pages declare 24 items while linking to 37 or 38 distinct prompt URLs each; its four category pages declare 20 and link to exactly 20.