Free · no account
What's holding your site back in Google and ChatGPT?
Paste your address. We fetch 25 pages, language versions included, and tell you exactly what's wrong and on which address, in plain English.
25 pages in about 3 secondsLanguage versions countNo email up front
Site scan
Working with the Neura MCP? Paste this one prompt into Claude or ChatGPT. That's all your AI agent needs to fix everything above.
The rest of the report in your inbox
Every finding, every address, plus what to do about it, in your inbox and right here in the report. Working with Neura MCP? You'll also get the ready-made fix prompt for your AI agent.
We keep your scan for 30 days. After that you'll get a short series of tips to make your site easier to find on your platform; unsubscribe with one click.Only visible from the inside
Four things no outside scanner can check
A crawler sees what Google sees. We also see what's behind it, because at Neuramerce the shop, the translations and the SEO are one system. These checks run automatically, every week, and fix what's unambiguous.
LanguageTranslation out of date?A source that changed while the translation stayed put. The storefront then quietly shows the old text.
status: stale · 14 fields · de
Routeshreflang cluster complete?Every language version must point to every other. That takes the slug register, not a crawl.
/en/about ↔ /de/ueber-uns ✓
CatalogueEAN, price, stock, feedProducts without an EAN, or with a price that can't be expressed as a round incl-VAT price. That means feed rejections before Google reports them.
4 products without EAN
RepairFixed automaticallyMissing descriptions are written in your brand voice and logged, so you can measure the effect. Everything else becomes a task with the reason attached.
13 descriptions → written
Join the waitlistBook 20 minutesFrom week one: scan on Monday, tasks on Tuesday.
What the scan measures
Six categories, seventeen checks: a complete technical SEO audit in miniature. Five categories any good scanner does; the sixth, language, takes a scanner that reads language versions as one cluster.
- Title & description
Search result
Title missing, too long or duplicated. Description missing, too short or duplicated, and in the language of the page, not just "identical". With a Google preview exactly as the searcher sees it. - Status & sitemap
Reachable & indexable
Status code per page, redirects with their destination, canonical present and self-referencing, noindex where it belongs. And: is every page in the sitemap, language versions included as alternates. - Headings & text
Content
Exactly one H1 per page, heading jumps (H1 → H3) that break the structure, and thin pages under 300 words. Eight pages with the same heading jump are usually one template, so one fix. - TTFB & images
Speed
Time to first byte per page, compression on the HTML, and images over 100 KB, with the file name, so you know which image to shrink. - og:image & llms.txt
Sharing & AI
og:title, og:description and og:image for share cards on WhatsApp and LinkedIn, Twitter Card, and llms.txt, so ChatGPT and Claude read your site the way you intended. - Only here
Language & hreflang
Does every language version point to every other, and back? Does html lang match the content? Is the description translated, or is there quietly a Dutch text on your English page? Outside scanners call that "duplicate". That's a different fault with a different fix.
From finding to fix
Not a list of 82 lines, but three things you do today. The scan groups by cause, not by address.
Sort by impact
Whatever sits in one template goes to the top. Eight pages with a heading jump are one fault. You see at a glance where one change fixes the most.See the address
Never "improve your meta descriptions". Instead: these three addresses, this text, this is what the searcher sees in Google right now.Apply it
On Neuramerce, automatically: missing descriptions are written in your brand voice and logged, so you can measure the effect. Anything that needs a decision becomes a task with the reason attached.
A crawler sees what Google sees. We also see what's behind it.
Compare3 models
Outside scanner
This site scan
On Neuramerce, every week
Description in the wrong language
Sees three identical texts and says "duplicate"
Untranslated, named per language
Fixed from the translation chain
Translation outdated after a source change
Invisible
Invisible
Reported per field and per language
hreflang cluster complete and reciprocal
Only what happens to be in the HTML
All 25 pages checked
From the slug register, for every page
Language versions in the sitemap
"Not in sitemap", often wrongly
Reads alternates in the sitemap
Generates them itself
Findings on a redirect
Measures the destination, reports the source
Skipped; status and destination shown
Route conflicts actively reported
EAN, price, stock, feed eligibility
No
No
Yes, before Google rejects the feed
Structured data per page type
Notes that it's missing
Notes that it's missing
Generates it
Repair
A fix plan in a call
Report with a fix per finding
Fixed or queued as a task, effect measured
Frequently asked questions about the site scan
Is the scan really free?
Yes. Score, categories, the main findings and the Google preview are visible right away, no account needed. The full report with all 25 pages goes to your inbox. That's what we ask your email address for, nothing else.
What is a technical SEO audit?
A check of everything Google needs before your content even counts: can the page load, is it in the sitemap, does it have one H1, a title and a description, does it point correctly to its language versions, does it load fast enough. This site scan is that audit in miniature: 25 pages, automated, in 3 seconds. It's not about whether your copy is good, but about whether Google and ChatGPT can take it in properly at all.
What is a good score?
Above 85 is healthy: the rest is maintenance. Below 70 there's usually one structural fault in a template that ripples across dozens of pages. You fix that once. Look less at the total and more at the category that stands out: a site can score 90 and still have pages showing the wrong language in Google.
How accurate is it?
We measure what's in the HTML, on 25 pages from your sitemap, language versions included. What we deliberately don't do: attach findings to a page that redirects, or call a language version "missing from the sitemap" when it's listed as an alternate. Those are the two noise sources we see most often in other scanners.
Why do language versions count?
Because that's where the fault sits most often and shows least. An English page with a Dutch description isn't broken. It just sits wrong in Google, and nobody notices. A scanner that only says "duplicate description" sends you to the wrong fix.
How often should I run an SEO audit?
After every bigger change: a new template, a new language, a migration. On Neuramerce you don't have to choose: the scan runs every Monday and the tasks are there on Tuesday.
Will this improve my Google ranking?
Not by itself. The audit removes what's holding you back; what you rank for after that is decided by your content. Anyone promising a top position based on a technical scan is promising something that can't be promised.
What happens to my data?
We keep your scan for 30 days, then it's gone. We use your email address for the report and a short series of tips for your platform; unsubscribe with one click, and we don't sell anything on.