Site images not being indexed

[webwonderment]

I have a website whose images where being indexed and ranked quite well in Google Images Search. Now since the end of last year in fact, images are not being indexed - only thumbnails. As far as I know, there's no technical reason why they aren't being indexed because it's the same as it's always been.

I thought about switching to a new domain - would this help? Like a fresh start. There's no warnings or penalties in Google Webmaster Tools so I haven't a clue why GoogleBot isn't indexing the big images.

[lucy24]

Preliminary questions: Are they being crawled? Not as often as pages, of course, but images do get re-crawled even if they've never changed. Do the images come up if you do an exact-match search? (Drag a copy of the actual image into the search box. Google is a bit capricious on "similar" images, but exact matches should always come up-- including different sizes of the otherwise-identical picture. Also a good way to find scrapers ;) )

[RedBar]

1. Are your images watermarked? If so many have been deleted or demoted so far no one will ever see them.

2. Goole's image SERPs is worse than its main search SERPs at the moment and is not improving, in fact it's a darned mess with many images ranking for queries that they're not even associated with.

3. Unless you have truly unique images with unique names it can take weeks, sometimes months, to even get them recognised.

I don't know whose idea it was at The Plex to do whatever they've done however they've made an absolute horlicks of what was a superb SERPs two or three years ago.

[aristotle]

Another possibility is that Google is ranking scraped copies ahead of your original images. Image theft is totally out of control.

[piatkow]

I gave up trying to make sense of rankings in image search years ago.

[RedBar]

Image ranking was ok until about three years ago, if you knew what you were doing, fine, however about that time Google told all webmasters to sharpen-up their images with alts, titles and even re-name them correctly...This was fine for a year and then Google decided to steal them all and make an absolute mess of their image SERPs.

Not only will scraped images outrank you but images with absolutely nothing whatsoever to do with one's specific subject/niche will outrank you.

They ought to admit defeat and go back to square one...yeah, as if that's ever going to happen.

[webwonderment]

GoogleBot visits pages and very rarely Google ImageBot grabs a big image. I think I've had 1 big image indexed in the past few months even though there's been 100's indexed.

Regarding hotlinking - Well there's loads of them hotlinked on other sites and they get the results. So I hotlink protected them including a no-cache so Google could hotlink but the cached image wouldnt appear on other sites. If I could figure a way out to know whether a Google image search hotlink is mine or refered from a scraper - I could hotlink protect the search results as well. It is very harsh on image webmasters who have worked hard, like myself, only to get hotlinked out of business. Imagine if search engines showed the entire page contents in it's standard search? There'd be an outcry. I don't mind a preview coming up in search, but the whole pic? Why would someone want to click-thru when they've got what they wanted? Great for users and search engines, disasterous for webmasters - you know, the ones that actually help build the web's content.

[webwonderment]

correction --> * even though there's been 100's added to the site.

[lucy24]

If I could figure a way out to know whether a Google image search hotlink is mine or refered from a scraper

When the new-style image search was first introduced, I did some hands-on experimenting and found that an image is actually requested twice: first for the lightbox thingie and then again if the user follows up by going to your site (whether page or image). But the first request isn't actually displayed; it's just held in reserve so it seems to load faster when the user clicks. What the user sees instead is a cached copy (most noticeable if your original was a high-resolution jpg, because the cached form is a png).

Normally a browser won't request the same image two times in a row when requests are only a minute or so apart; that's what a browser cache is for. So most of those requested files are never even seen by the human user. You can reduce drag on your site by rewriting the first one (referer in "blank.html") to, say, a one-pixel gif and flagging this gif to inspire instantly. The browser then only requests your real file if the user actually goes there. And if it was a hotlink, the request will now come through with the offending site as referer, so you can easily deal with it in the usual way.

[webwonderment]

Thanks for the feedback it's much appreciated.

1. Are your images watermarked? If so many have been deleted or demoted so far no one will ever see them.

No, none.

Another possibility is that Google is ranking scraped copies ahead of your original images. Image theft is totally out of control.

Not only will scraped images outrank you but images with absolutely nothing whatsoever to do with one's specific subject/niche will outrank you.

Exactly what is happening. Add to the fact they aren't even indexing mine kind of rubs salt in the wound.

I did some hands-on experimenting and found that an image is actually requested twice: first for the lightbox thingie and then again if the user follows up by going to your site (whether page or image).

That depends if you set the Expires:0 or No-cache header on the image. If set, the image is requested twice, if not it's the cached version they see.

Normally a browser won't request the same image two times in a row when requests are only a minute or so apart; that's what a browser cache is for.

I've set all the images to expire immediately so there should be no browser caching at all. This means that a scraper who hotlinks and gets the image onto Google will have a dummy pic on their page - so it makes it pointless for them to put the images on - people only see a dummy image.

I've got a whitelist of sites that can hotlink - mainly search engines and sites that link back. To get traffic I've realised there's little choice other than to let the search engines hotlink the images because I doubt they'd keep them in the index long if it's just a 1 pixel gif that's shown in their search lightbox. It sucks but there's little choice.

Anyway regarding the lack of indexing - I've added images to the GWT sitemap - we'll see if that helps.

[lucy24]

I've set all the images to expire immediately so there should be no browser caching at all.

In my version, your server only has to cough up a few bytes for the single pixel. (On my system it's something like 372, give or take.) Since you already know the user won't be seeing the image, there's no point in sending out the whole 2, 20 or 100+ kilobytes of the real thing.

[RedBar]

I'm sorry to say that this is the catastrophic new Google image world, personally I have removed tens of thousands of images because of Google's theft making it almost impossible for anyone to derive any income from their own hard work.

Hotlinking is easy to resolve, stopping all and sundry copying without watermarks is a nightmare. These days I don't serve as many images in 24 hours as I used to in a few minutes.

Whomsoever's idea it was at Google to introduce this really arrived with no brains that day and it certainly has not benefitted the user/consumer one iota insofar as my widgets are concerned, it's a step back almost 20 years!

[webwonderment]

Whomsoever's idea it was at Google to introduce this really arrived with no brains that day and it certainly has not benefitted the user/consumer one iota insofar as my widgets are concerned, it's a step back almost 20 years!

Google, Bing & Yahoo all do it though. Whoever did it first has made the others do it I imagine.

There's tons of hotlinks and stolen pics on that pic match tool thing on Google. What can I do? I'm seriously getting upset with it now. Putting stolen copies and uncredited hotlinks of my hard work way above mine? It's like I've been dropped to the bottom and thieves are benefitting. What the hell is going on?

[webwonderment]

Google know that mine are the original because they were retrieved first. How can copies supersede originals?

[webwonderment]

I've just seen a stolen pic that's 2nd ranked in Google, that's got a watermark hotlink protection on it! That's right - they've stolen my work and hotlink protected it. This is an absolute disgrace - I'm fuming.