← All case studies
Live internal SEO audit & website fixes · Squarespace, Google Search Console

Invisible on Google — and quietly broken underneath.
Here's what fixing both actually took.

A real, currently-operating small business's Squarespace website — live for years, with real clients and real testimonials — had never once shown up in a Google search of its own domain. This page is the full Squarespace SEO audit and the website fixes that came out of it: missing meta descriptions, unverified Search Console, broken structured data, and inconsistent contact info, plus the two places a fix looked done and wasn't — a platform bug that silently ate a metadata save, and a connector limitation that forced a manual pivot instead of an automated one.

Verified live — SEO audit and website fixes shipped 2026-07-30, two issues still open
CASE STUDY
THE PROBLEM

A site can be live, staffed, and serving real clients — and still not exist on Google.

Running site:digitizeservices.net returned zero results. Not page two, not a thin index — zero. Google Search Console had never been verified for the domain, so there was no ownership signal, no submitted sitemap, and no visibility into how Google saw the site at all. A business can do everything right on the page — real testimonials, real client logos, a working contact page — and still be functionally invisible to anyone searching for it, because none of that matters if the search engine was never told the site exists. Digging further turned up the same story below the surface: meta descriptions missing, structured data incomplete, contact details inconsistent across two similar domains. An SEO problem and a website-maintenance problem, tangled together.
CASE STUDY
WHAT CHANGED

From zero search presence to a verified, submitted, partially-fixed site.

No Google Search Console verification — no ownership signal, no index data, no way to even see the problemVerified via an HTML meta tag injected through Squarespace's header code injection; sitemap.xml submitted the same session
A key service page (/pawnshops) had an empty meta description — nothing for Google to show in a result snippetWritten and confirmed live
Organization and LocalBusiness structured data had no business address at allAdded — using the address Eric confirmed directly, not an unverified one that turned up in a web search for the business name
Two near-identical domains (hyphenated and not) both in live use as a contact email, inconsistently, across the site and this repo's own docsStandardized on the one that matches the live domain, updated everywhere it appeared — site identity, JSON-LD, footer mailto link, contact page, internal docs
56 of 58 homepage images had no alt text; ~48 client-logo images unidentifiable to a screen reader or an image searchPartially fixed — see the pivot below; the rest needs manual entry, and that's stated here rather than smoothed over
CASE STUDY
THE STRUGGLE

Three fixes that looked done and weren't — for three different reasons.

Most of this audit went the way a checklist suggests it should: find a gap, write the fix, confirm it's live. Three items didn't, and all three are worth walking through — because "the tool said success" turned out to mean something different each time, and each one needed a different fix: a platform bug logged as open, a broken tool path replaced with a manual one, and a website edit re-checked twice instead of once.

SEO fix — still open

The homepage's SEO title and meta description were rewritten through Squarespace's page-metadata API. The call reported success: true. The Squarespace editor's own SEO tab showed the new text saved correctly — confirmed directly by Eric, not assumed. But the live, publicly rendered homepage kept serving the old <title> and <meta name="description">, unchanged, tested repeatedly over several minutes. A distinct placeholder value was tried specifically to rule out "no diff detected" silently skipping the write. The response's age: 0 header ruled out a stale CDN cache too. The save reached Squarespace and reached the editor's own display of the saved state — and still never reached the rendered page. That's a gap between two things Squarespace itself reports as true, and it's on their side to close, not something a retry or a cache-bust fixes.

SEO fix — pivoted, not solved

Every section on this site's pages uses Squarespace's newer Fluid Engine layout. The image-editing tool identifies a target block by matching its existing title, description, or alt text — which works fine on Squarespace's older layout, and fails outright on Fluid Engine. Confirmed by testing against a block with real, human-written alt text already in place ("eric bagnes" on the About page photo) — still returned "no block found." That's not a one-off miss, it's a hard incompatibility: this class of tool cannot write to any image block on a Fluid Engine site, full stop.

Website fix — caught by re-checking, not assumed

Standardizing the contact email meant editing a mailto link embedded inside a Fluid Engine text block — the kind of small, routine website fix that's easy to consider done after one pass. It wasn't: on this layout, a single text-block update call was found to sometimes change only the link's href or only its visible label, not reliably both together in the same call. Nothing in the response flagged which half had actually changed. The fix was procedural rather than code — treat the href and the visible text as two separate things to verify, and always re-check both after any mailto edit, since a second identical-looking call usually catches whichever half the first one missed.

All three incidents got the same response: stop retrying a path that's confirmed not to work, and say so plainly instead of quietly downgrading the claim. The metadata bug is logged as open, waiting on a platform-side fix. The alt-text gap became a manual task instead of an automated one — filename-based suggestions were drafted for images with descriptive names ("western logo.jpeg" → Western Healthcare Nursing Services), and the rest were flagged as needing a person to actually look at the image, because guessing at alt text from nothing is worse than leaving it blank. And the mailto bug turned into a standing rule for every future edit on this site: never trust a single call on a two-part field, verify both halves every time.
CASE STUDY
HOW IT'S BUILT

Verification over assumption, on every SEO and website claim alike.

The same principle behind every other system on this page applied to the audit itself: don't report a fix as done because a tool call returned success, confirm it against the actual live, public-facing result. That's what caught the metadata bug — a naive audit would have logged the title fix as complete and moved on, because every signal short of the rendered page said it worked. It's the same discipline that caught the mailto half-update, and the same reason the alt-text gap wasn't quietly skipped: once a tool path was confirmed broken on this layout, the honest move was a manual, human-executable fallback, not silence.

Two smaller findings resolved themselves without any fix at all — a ghost sitemap entry and a page missing from the sitemap both self-corrected during a routine sitemap regeneration cycle later in the same session. Not every anomaly needs an intervention; some need a second look after the platform's own housekeeping catches up. The audit noted both rather than claiming credit for a fix that wasn't one.

Wondering if your own site actually shows up?

"It's live" and "it's indexed" are two different claims. Happy to run the same kind of SEO audit and website check against your own site and tell you honestly what's actually true today.