How does probeo help you hold an agency accountable?

The short answer

It gives you an independent read of the site your agency is paid to maintain, so you can see what has and has not been done. The conversation stops being their word against yours.

A mid-market company used it when it was paying an agency every month for articles and maintenance and could not get it to act.

Paying every month, and nothing moving

They paid an agency a monthly retainer for articles and site maintenance. Asking what had been done got answers, and the site did not change.

Without a way to check the site yourself, you cannot tell a slow agency from a hard problem.

What they could not see

  • What was actually wrong with the site, apart from the agency's own reports.
  • Whether the problems the agency was paid to fix were getting fixed.
  • Whether the articles being paid for were doing anything.

What probeo showed them

probeo read the whole site and listed what was wrong, page by page, in order of what it was costing them. None of it depended on the agency.

What they did with it

They let the agency go. Then they rebuilt the site and fixed everything probeo had found.

Where it stands

The site is rebuilt and the issues are fixed. The content is next: they have moved on to drafting and rewriting pages, and are waiting on the first of them.

What you can check yourself

Each of these is independently verifiable, and none of them requires anyone's cooperation.

Read the traffic reports carefully, though. Google omits low-volume queries to protect privacy while still counting them in the chart, stores only the most important rows, covers only a subset of a site's URLs, and delivers data in two to three days with a thousand-row limit.123456

  • Whether a change reached Google at all. URL inspection reports the most recently indexed version of a page and separately fetches the live one, so the two can be compared.
  • What Google actually holds. The same tool shows the last crawl date, the raw HTML Google received, and both the canonical you declared and the one Google selected.
  • Whether work was accepted. Pages Google fetched and declined to index are reported as crawled and currently not indexed, and Google says submitting a URL does not guarantee it appears in the index.
  • Whether structured data is valid. The rich results reports show valid and invalid items and let you track whether fixes worked.
  • What the site looked like before. The Wayback Machine stores captures by date, though it warns that not every date for every site is complete and that pages blocked by robots.txt may be missing.

What an unmaintained site looks like

These are the states that accumulate when nobody is watching, and each is documented rather than a matter of taste.

A maintained site is not one with no findings. It is one where the findings are known and the record shows them being worked through.789101112

  • Robots rules used as a hiding place. Google says robots.txt is not a mechanism for keeping a page out of Search, and that a disallowed page can still be indexed if other sites link to it.
  • Sitemaps that lie. Google uses the last modified date only if it is consistently and verifiably accurate, defines a significant update as a change to main content, structured data or links, and ignores priority and change frequency entirely.
  • Canonicals that contradict each other. Declaring one URL in a sitemap and a different one in a link tag for the same page is named as something not to do.
  • Errors left to settle. Repeated 5xx responses make Google slow its crawling and eventually drop indexed URLs, while soft 404s are error pages answering with a success code.
  • Structured data that describes something the page does not show, which can cost the page its rich result through a manual action.
  • Security issues. Hacked content placed through a vulnerability can put a warning label on your results, and review after a fix takes days or weeks.

What this cannot tell you

probeo shows what state the site is in. It does not tell you why an agency did or did not do the work, and it is not a verdict on whether they were worth the fee. It gives you the facts to have that conversation.

It cannot tell you who did what. No primary source records authorship, and Search Console has no history of which URLs a given user submitted or inspected.

It also cannot settle whether an outcome was caused by the work. Google says core updates are broad and do not target specific sites, recommends waiting at least a week after one completes before analyzing, and says recovery can take months.13

And speed is not the lever it is often sold as: Google states there is no single page experience signal and that good Core Web Vitals do not guarantee rankings.14

Questions

Does the agency have to know?

No. probeo reads the public site. Nothing has to be installed and nobody has to give it access.

What counts as a page?

A managed page is any page you want probeo to look at. You choose which.

How do I know whether a page was actually changed?

Inspect the URL. Google shows the last crawl date and the HTML it holds, and a live test fetches the current page, so you can see whether what is live matches what Google has and when it last looked.

New pages went up but nothing happened. Why?

One documented possibility is that Google crawled them and chose not to index them, which it reports as crawled and currently not indexed. Submitting a URL does not guarantee indexing.

Why do the agency's numbers differ from Search Console?

Search Console omits low-volume queries for privacy, keeps only the most important rows, covers a subset of URLs and caps tables at a thousand rows. Two honest reports can disagree for documented reasons.

They say the site is accessible. Can I verify that?

Only partly. Conformance is defined for full pages and cannot be claimed if part of a page is excluded, and W3C states that evaluation tools cannot determine accessibility and that human judgment is required.

Give us a URL. No credentials, no integration, nothing behind a login. We walk you through what we found on your own pages.

See what AI says about you

Everything this page is built on

  1. https://support.google.com/webmasters/answer/9012289
  2. https://support.google.com/webmasters/answer/7440203
  3. https://support.google.com/webmasters/answer/7552505
  4. https://help.archive.org/help/using-the-wayback-machine/
  5. https://support.google.com/webmasters/answer/17011259
  6. https://support.google.com/webmasters/answer/96568
  7. https://developers.google.com/search/docs/crawling-indexing/robots/intro
  8. https://developers.google.com/search/docs/crawling-indexing/sitemaps/build-sitemap
  9. https://developers.google.com/search/docs/crawling-indexing/consolidate-duplicate-urls
  10. https://developers.google.com/crawling/docs/troubleshooting/http-status-codes
  11. https://developers.google.com/search/docs/appearance/structured-data/sd-policies
  12. https://support.google.com/webmasters/answer/9044101
  13. https://developers.google.com/search/docs/appearance/core-updates
  14. https://developers.google.com/search/docs/appearance/page-experience