SaaS site boosts AI-driven search traffic

The team behind a LATAM SaaS platform rolled out a four-phase SEO overhaul on June 5, 2026, adding structured data, explicit AI-bot directives, and a lightweight pre-render pipeline [DevTo].

── What shipped ──

• PageMeta extension – The React-Helmet wrapper now accepts ogType and jsonLd props, emitting Open Graph tags and up to six <script type="application/ld+json"> blocks per page [Google Rich Results].
• Schema constructors – A tiny schema.ts library provides articleSchema, breadcrumbSchema, personSchema, etc. Six schema types are now rendered on every relevant route.
• hreflang – Every render includes hreflang="es-mx" and an x-default tag, eliminating a 5-10% silent index loss observed on multilingual queries.
• llms.txt – A new llms.txt file at the site root follows the 2026 specification and lists private paths (/dashboard/*, /admin/*) as Disallow. Robots.txt now contains explicit Allow blocks for GPTBot, ClaudeBot, PerplexityBot, Google-Extended, CCBot, and Anthropic-AI.
• Dynamic sitemap index – The static sitemap.xml was replaced by a <sitemapindex> referencing three dynamic sitemaps for blog, profiles, and salary reports. The backend endpoint /sitemap-blog.xml caps each map at 50,000 URLs, matching Google’s limit.
• IndexNow integration – After each post publish, the service fires an IndexNow POST to api.indexnow.org with up to 100 URLs [DevTo].
• Web-Vitals sampling – The front-end reports CLS, INP, LCP, FCP, and TTFB at a 10% sample rate to CloudWatch. An alarm triggers on p75(LCP) > 3000 ms.
• Static HTML snapshots – A 100-line post-build script patches dist/index.html for eight landing routes, inserting title, description, OG, and canonical tags.

The team added JSON-LD schema markup, which directly shapes the snippets shown to users [Google Rich Results]. IndexNow integration notifies Bing, Yandex, Naver, and Cloudflare instantly, reducing discovery latency. Static HTML snapshots also slash build cost by removing Puppeteer, saving roughly $200 per month in CI compute.

The llms.txt file protects both compliance and SEO performance by respecting explicit crawl directives [DevTo]. Sites that omit llms.txt risk having private surfaces indexed and inadvertently feeding proprietary data to LLMs.