Fetched as plain HTML, every Soria page (/, /plans, /careers, /support, /trust and /changelog) returns the same title, "Soria - Superintelligent investment research", and no body text. The headline, the proof and the demo button only appear after JavaScript runs, and the sitemap lists five URLs, leaving out /plans, /trust and /changelog. Soria sells to analysts who now research vendors through search and AI assistants, and it ships its own assistant connector. Yet a reader that fetches soriaanalytics.com finds a tagline and nothing else. Soria's strongest claims (clients include 5 of the top 10 investment banks and 8 of the 20 largest hedge funds; coverage built from SEC, CMS, IR and state sources) appear only on its YC page and third-party directories, one of which calls the public site "a single page".
First action
Open view-source on the homepage and /plans and list every headline, claim and button that is missing from the raw HTML.
How to measure it
Pages readable without JavaScript
- Now
- –
- Target
- All 8 public pages
- By
- 2 weeks
Steps
Confirm the gap
Plain fetch of the six linked pages; record title, description and body text for each.
Co-founder & CTO 1 hour
Choose the rendering
Static generation for marketing routes; the logged-in app stays as it is.
Co-founder & CTO 1 day
Write page metadata
A unique title and description for home, plans, trust, changelog, careers and support.
Founder & CEO 2 hours
Put the proof in the HTML
Client claims, source types and delivery channels as text, not only images or animation.
Founder & CEO half a day
Fix the sitemap
Add /plans, /trust and /changelog with real last-modified dates.
Founding Engineer 1 hour
Add structured data
Organization and SoftwareApplication markup with the same facts.
Founding Engineer 2 hours
Verify
Search Console URL inspection, a link-preview test in Slack and email, and a repeat plain fetch.
Co-founder & CTO 1 week
Why it works
Pre-render the marketing pages at build time (static generation or server rendering on the current setup), so each URL ships its own title, meta description, H1, body copy and demo link in the HTML. The app stays client-side. Give /plans, /trust, /changelog and /careers distinct titles and descriptions, add them to the sitemap with real last-modified dates, and add Organization and SoftwareApplication structured data that repeats the proof points as text. Check the result with a no-JavaScript fetch, Search Console URL inspection and a link-preview test. Speed is already excellent (98 on mobile, and real visitors pass Core Web Vitals), so this is about being readable, not faster. The upside: Soria can be found and described correctly when an analyst searches or asks an assistant for "AI research terminal for healthcare investors", and a link pasted into a chat with a PM shows a real preview.
Many crawlers and assistant fetchers read the first HTML response and run little or no JavaScript. A page that ships its content in the HTML gets indexed, quoted and previewed correctly. The robots.txt already allows search and assistant fetchers and only blocks training crawlers, so the empty HTML is the only thing in the way. For a team of three, it is a one-week change that works on every future visit.
Ready to use
ChecklistReadable-site checklist
[ ] Homepage HTML contains H1, subhead, proof and demo link [ ] Unique <title> and meta description on /, /plans, /trust, /changelog, /careers, /support [ ] Sitemap lists every public page with real lastmod [ ] Organization + SoftwareApplication JSON-LD [ ] Open Graph title, description and image per page [ ] Plain fetch (no JavaScript) shows the copy [ ] Search Console URL inspection passes [ ] Link preview checked in Slack and email
Risks and how to handle them
Pre-rendering breaks interactive parts of the homepage
Render the text server-side and hydrate the interactive pieces after load.
Claims in public text need sign-off from clients
Use counts and categories ("5 of the top 10 investment banks") rather than names.
Evidence
- Your websiteHomepage and inner pages fetched as plain HTML (November 10, 2026): /, /plans, /careers, /support, /trust and /changelog each return the same title, "Soria - Superintelligent investment research", and no body text; the headline, proof and calls to action appear only after JavaScript runs. Meta description: "Superintelligent investment research. Analyst-grade research and data terminal, starting with healthcare."Source
- Your websiteSitemap (checked November 10, 2026): 5 URLs (/, /careers, /support, /privacy, /terms), all last modified August 28, 2026; it leaves out the /plans, /trust and /changelog pages the site links to. No blog, case studies, newsletter or published prices were found; demos are booked through a Cal.com scheduler on the homepage, and the only social link is LinkedIn.Source
- PageSpeedGoogle PageSpeed (November 10, 2026): mobile performance 98, desktop 100, SEO 100, accessibility 96 (one colour-contrast issue); the largest element is the H1 "Superintelligent investment research". Real visitors pass Core Web Vitals: LCP 982 ms, INP 63 ms, CLS 0.01 at the 75th percentile. Google Tag Manager is the main third-party weight (178 KB).Source
- Your websiterobots.txt: open to all crawlers by default, with 32 named AI-training crawlers disallowed and an "ai-train=no" content signal; search engines and assistant search fetchers are not blocked by name, so what limits them is the empty HTML, not the rules.Source
- HiringFounding Engineer (New York, $150K-$175K, 2.5%-5% equity, new grads welcome): ship customer-facing features with real-time updates, build pipelines and tools so internal analysts can add sources, improve the agent harness, and within 30 days build the company ontology and ship the first version of an "Earnings Center". The post says clients include "5 of the top 10 investment banks and 8 of the 20 largest hedge funds".Source
- Directory listingsFintech Sandbox calls Soria "the only AI-native 'Bloomberg terminal' built specifically for sector specialists", built from hundreds of sources including federal and state agencies, nonprofits and private vendors. Huntscreens lists IR materials, SEC filings, CMS data and state portals, dashboards that end Excel tracking, and alerts by email, UI, API, MCP or warehouse sync. apis.io describes the public site as a single page, with pricing and trial unknown.Source