Google reverse image search, and what it will not do for you

The six things the search does not do

1. Confirm the copy is still there

Index entries lag reality. Open each page and confirm the image is on it now. A notice filed over content that has already gone is rejected, and rejections are recorded.

2. Establish that you own it

Harder than it sounds, and the most common reason a business files something it should not have. Embedded stock, a theme’s demo images, a freelancer’s photographs with no written assignment — in each case the person filing may not be the rights holder, and swearing otherwise under penalty of perjury is a serious thing to get wrong.

3. Consider fair use

Lenz v. Universal requires a copyright holder to consider whether a use is fair before sending a notice. Commentary, criticism, news reporting and parody all exist, and a verbatim copy on a competitor’s sales page is a very different case from a screenshot in a review.

4. Identify the host and its designated agent

A notice goes to the host’s registered copyright agent, not to a support inbox. Finding that agent means resolving the domain, identifying the IP owner, and searching the US Copyright Office directory — and checking the designation has not expired, because they lapse after three years.

5. Write a compliant notice

Six elements under § 512(c)(3)(A), two of them sworn statements that should not be paraphrased. Specific URLs, not a homepage.

6. File it the way that recipient accepts

The routes genuinely differ. One major host will not open attachments at all and needs everything in the plain-text body. One demands an uploaded letter of authorization before it will act for an agent. Several forward your entire notice, contact details included, to the person who copied you.

De-indexing is not removal

This is the misunderstanding that costs people the most time, so it is worth being blunt.

Google will accept a copyright removal request and remove a URL from its search results. The page itself stays exactly where it is, on the host, reachable by anyone with the link, and still visible in every other search engine. You have removed a signpost, not the thing it pointed at.

There are cases where de-indexing is genuinely the goal — if the harm is that the copy outranks you, taking it out of results fixes the harm. But if a competitor is using your photographs to sell a product, de-indexing changes nothing about that. The notice has to go to the host.

One report per product

A detail that catches people out: Google treats its products separately for this purpose. A copy on a Blogger site that also appears in Search and on YouTube is three reports, not one. There is no single filing that covers Google as a whole.

And Google Search is only one search engine. Bing runs its own copyright removal process through Webmaster Tools, and a de-index at one does nothing at the other.

Where we come in

Everything above can be done by hand, and for a single copy of a single image we would encourage you to do exactly that. Our notice builder is free, it produces a notice containing all six elements, and it sends nothing anywhere — the text is yours to use however you like, including without us.

What we are for is the case where that becomes unreasonable: a whole product catalogue lifted, a site cloned across dozens of pages, copies spread over a host and a marketplace and a search engine at once, each with its own route and its own requirements. We scan your site for both words and images, verify each match against the live page, identify the host and the designated agent for every one, consider fair use and record that consideration, file as your authorised agent, and re-check afterwards. You are quoted per URL in advance and billed only for pages that come down.

We cannot promise removal and we do not. Hosts and search engines decide. What we can do is make sure the notice reaches the right person in the form they accept, which is most of what determines whether anything happens at all.