SEO Crystal Ball Case Study
This case study follows a real Your SEO Wizard site review: initial scan, source changes, routing diagnosis, redeploy, and repeated verification.
Follow-up scan
The site already had a strong baseline. The real value came from using repeated findings to choose specific fixes, then checking whether those findings cleared after deployment.
A+ site grade
Whole-site review
Warnings cleared
Warnings cleared
What changed
Crystal Ball flagged missing JSON-LD on pricing, SEO Roadmap, SEO Crystal Ball, and SEO Graph Wizard. It also found short or weak titles on privacy, terms, and app pages.
The structured-data and title findings cleared after source updates and redeploy. A later routing inconsistency was traced beyond page content, corrected at the host and rewrite layers, and converted into a repeatable deployment check.
Representative fixes
Initial finding: JSON-LD structured data was missing.
Action: Added WebPage, BreadcrumbList, and OfferCatalog schema aligned with visible pricing content.
Follow-up: Structured-data warning cleared.
Initial finding: app pages had short titles and missing structured data.
Action: Expanded titles and added SoftwareApplication plus breadcrumb schema for Crystal Ball and Graph Wizard.
Follow-up: Title and JSON-LD warnings cleared.
Initial finding: privacy and terms pages had titles that were too thin for clear identification.
Action: Rewrote titles as Privacy Policy | Your SEO Wizard and Terms of Service | Your SEO Wizard.
Follow-up: Title-related warnings cleared.
Technical diagnosis
After the schema and title fixes cleared, repeated browser-like requests exposed intermittent status behavior on key URLs and redirect problems on education routes. That changed the work from content editing to host-level diagnosis: canonical URL settings, clean extensionless routes, redirect hops, document-root behavior, and physical directories that conflicted with flat-file rewrites.
A clean public URL should resolve predictably. Inconsistent status responses can confuse crawlers, weaken canonical signals, and make scans harder to interpret.
The fix combined one host-level HTTPS canonical setting, explicit flat-file routes, removal of a conflicting physical education directory, and repeated browser-like GET checks. A strict public deployment allowlist now prevents source and workspace folders from reaching production.
Durable lessons
Not every warning requires a page edit. The useful outcome is a process that separates content, metadata, routing, deployment, and scanner-policy decisions before recommending a fix.
Test important URLs more than once with browser-like GET requests so intermittent status failures are not hidden by a single successful response.
Build a clean deployment artifact containing only public pages, required API files, and referenced assets instead of uploading the working directory.
Use Article, SoftwareApplication, OfferCatalog, Service, FAQPage, and breadcrumb markup only where visible page content supports it.
Distinguish confirmed defects from contextual recommendations so a technically healthy page is not scored like a broken one.
Repeat the workflow
The tutorial turns this case study into a practical checklist for reviewing any public page or small website.