Configure search, social sharing, and AI visibility

Give every public page one identity #

• Title: a concise, specific description of this page, not the same site-wide phrase everywhere.
• Description: a useful summary for people deciding whether to visit.
• URL: short, readable and stable.
• Canonical: the one public address that should represent duplicate variants.
• Index state: intentionally index or exclude the page; do not leave private or unfinished content discoverable.

Fronty hosting emits baseline metadata, Open Graph/Twitter tags, a WebSite structured-data block, robots and a sitemap. Readiness can detect missing or placeholder values, but you still supply accurate page-specific content.

Configure item pages and social previews #

Bind a dynamic item's title, description and image to the appropriate table fields. Open at least two item URLs and inspect their rendered source to confirm that metadata changes with the record rather than repeating the template page.

Choose a social image whose content remains understandable when cropped. Fronty does not currently provide rich schema selection for every business type, and its generic WebSite schema is not a substitute for Product, Article, Event or LocalBusiness data where those are accurate and supported.

Understand discovery limits #

A sitemap helps crawlers discover listed URLs; it does not guarantee indexing. Current generated-site sitemaps cover normal live pages but do not enumerate every parameterized item route. Link important items from crawlable pages.

Multilingual presentation exists, but automatic hreflang output and immutable translation revisions are not complete. Do not claim a locale is fully indexed until its public URLs, canonicals, language signals and release behavior have been verified.

Provide information for AI agents honestly #

Fronty's marketing site publishes an /llms.txt index. Customer sites do not currently receive a dedicated per-site llms.txt automatically. If that file matters to your publishing strategy, add and maintain it in the host you control after export or wait for explicit product support.

AI-agent files are discovery aids, not access control and not a guarantee that an assistant will cite or summarize the site. Keep important claims in accessible HTML, use stable links, and expose only public information.

Know what the built-in analytics can prove #

Built-in analytics currently reports a 30-day page-view total, a daily page-view series and the ten most-viewed paths. The collection endpoint is cookieless and respects Do Not Track and Global Privacy Control signals.

That view does not currently establish unique visitors, acquisition sources, devices, funnels, conversions, form completion, checkout performance or retention. Use the numbers as a compact traffic view, not as a complete growth analytics system. If a business decision depends on one of those missing dimensions, connect an appropriate analytics service and document its consent and privacy behavior.

Verify the rendered public result #

1. Open the public page source and confirm one title, description and canonical.
2. Check Open Graph and Twitter values against the page, including an absolute image URL.
3. Open /robots.txt and /sitemap.xml on the public hostname.
4. Inspect two dynamic items if the site uses item pages.
5. Validate structured data and request indexing through the search tools you use.