Google Image Search - Can I Control My Images G displays?

[darren13]

Hello All,

To get straight to the point:-
We have 200 or so images which Google shows on it's Image Search (http://images.google.co.uk) results page when I search for New Forest Holiday Cottage (New Forest Holiday Cottages (http://images.google.co.uk/images?hl=en&q=new+forest+holiday+cottage&btnG=Search+Images)).

What I would like to do is tell Google which images (maybe 5 or 10) I would prefer it to show first. Now I know there isn't a hotline for this sort of thing at Google! So was wondering if there was a tag of some kind which would allow me to identify to Google images which I deemed to be of more importance?

Any help as ever is greatly appreciated guys.

All the best,
Darren.

[stymiee]

Unfortunately there isn't. Their image search results are algorithmic just like their content search. You can influence their Image SERPs just like you can influence their content search (i.e. SEO for images) but that won;t allow you any real control over what order they show images.

[spenland]

I wonder what would happen if you placed images you didn't want indexed into the robots.txt file? Never tried it, but just curious.

[briguy]

how about the "no follow tag" in your img src ? or changing your "alt tag" to some other keywords maybe your domain name?

[kgun]

Disallow:
User-agent: Googlebot-Image

[Peter (IMC)]

SEO is SEO, also for images. Just try to imagine what factors are important to a search engine when determining ranking of the images.

Likely factors:

* Image size
* Alt text
* File name
* Surrounding text

Notice how the first image in the results is shown on your website as a small image, while in fact the real image is big, but scaled down by the HTML code. Google however shows the real size of the image in the result. I wouldn't be surprised that if you resize that image to the size you want it to be shown in the site, it drops in rankings a bit. That also would considerably reduce your traffic costs and download time of the page. So in any way a good idea to do.

Play with the above factors and see what happens. My best bet right now is that if you change the file names of those images you want higher up into the search phrase, you'll some nice results.

[Webnauts]

Darren create a second folder with the images you want to have displayed in Google, and disallow the indexing of your existing images folder in your robots.txt

[magic_majax]

According to this page (http://www.linktutorial.com/read/articles/attribute-relnofollow) the img tag can take a nofollow attribute.

Just build and test using the w3 valiadator (http://validator.w3.org/) if you don't want to break links in the existing google index (I don't think google is overly keen on this)

[Webnauts]

According to this page (http://www.linktutorial.com/read/articles/attribute-relnofollow) the img tag can take a nofollow attribute.

Just build and test using the w3 valiadator (http://validator.w3.org/) if you don't want to break links in the existing google index (I don't think google is overly keen on this)

Does that mean that using the nofollow attribute, the images will not be indexed? I never heard that before. Very strange. It is though to avoid passing values, and not to hinder indexing.

[amar]

We don't have option to display images in priority. However we can restrict the unwanted images by placing them in a separate folder and restricting them using a robots.txt file.

Thanks,
Amar


SEO India (http://amarseo.blogspot.com)

[garywhw]

Amar, that's smart.

[Webnauts]

Amar, that's smart.
Did I miss something here? Didn't I mention that before?

[Webnauts]

Disallow:
User-agent: Googlebot-Image

Kgun, if we use that, then all images will be excluded. I think you missed the thread starters question.

[Webnauts]

I agree with Peter (IMC) post above, but I think you must consider using images the following:

An alt attribute is not a label or description for the image. This is not an immediately obvious distinction. In fact, it might seem natural to assume that alternate text is a label or a description. It is not. The words used should be a text equivalent and convey the same information/serve the same purpose as the image. The aim is to provide the same functional information that the visual user sees. The alt text should be a "stand in" if you will if the image is missing. The test is, when you replace the image with the text would everyone do the same thing/receive the same information as best as possible?

Some examples:
- If a search button is a magnifying glass or binoculars its alt text should be 'search' or 'find' not 'magnifying glass' or 'binoculars'.
- If an image is meant to convey an image - then a description is appropriate.
- If it is meant to convey data - then that is what is appropriate.
- If it is meant to convey a function then that is what should be there.

Some Alt Attribute Guidelines:
- Always add alt attributes to images. Alt is mandatory for accessibility and for valid (X)HTML.
- For images that play only a decorative role in the page use an empty alt (i.e. alt="") or a CSS background image so that reading browsers do not bother users by uttering things like "spacer image".
- Remember it is the function of the image we are trying to convey. For instance any button images should not include the word "button" in the alt text. They should emphasize the action performed by the button.
- Alt text should be determined by context. The same image in a different context may need drastically different alt text.
- Try to flow alt text with the rest of the text because that is how it will be read with adaptive technologies like screen readers. Someone listening to your page should hardly be aware that there is a graphic there.

I've found it useful categorize non-text content into three levels:

1. Eye-Candy:
Eye-Candy are things that serve no purpose other than to make a site visually appealing/attractive and (in many cases) satisfy the marketing departments. There is no content value (though there may be value to a sighted user). Never alt-ify eye-candy unless there is something there which will enhance the usability of the site for someone using a non-visual user agent. Use a null alt attribute or background images in CSS for eye-candy.

2. Mood-Setting:
This is the middle layer of graphics which may serve to set the mood or set the stage as it were. These graphics are not direct content and may not be considered essential, but they are important in that they help frame what is going on. Try to alt-ify the second group as makes sense and is relevant. There may be times when doing so may be annoying or detrimental to other users. Then try to avoid it. For instance ALT text that is identical to an adjacent text is unnecessary, and an irritant to screen reader users. I recommend alt="" or background CSS images in such cases. But sometimes, it's important to get this content in there for all users. Most times it depends on context. The same image in a different context may need drastically different alt text. Obviously, content should always be fully available. The way you go in this case is a judgement call.

3. Content/Function:
This is where the image is the actual content. Always alt-ify content and functional images. Title and long description attributes may also be in order.

The reason many authors can't figure out why their alt text isn't working is that they don't know why the images are there. You need to figured out exactly what function an image serves. Think about what it is about the image that's important to the page's intended audience. Every graphic has a reason for being on that page: because it either enhances the theme/ mood/ atmosphere or it is critical to what the page is trying to explain. Knowing what the image is for makes alt text easier to write. And practice writing them definitely helps.

A way to check the usefulness of alternative text is to imagine reading the page over the telephone to someone. What would you say when encountering a particular image to make the page understandable to the listener?

Besides the alt attribute you have a couple more tools at your disposal for images.

First, in degree of descriptiveness title is in between alt and longdesc. It adds useful information and can add flavor. The title attribute is optionally rendered by the user agent. Remember they are invisible and not shown as a "tooltip" when focus is received via the keyboard. (So much for device independence). So use the title attribute only for advisory information.

Second, the longdesc attribute points to the URL of a full description of an image. If the information contained in an image is important to the meaning of the page (i.e. some important content would be lost if the image was removed), a longer description than the "alt" attribute can reasonably display should be used. It can provide for rich, expressive documentation of a visual image. It should be used when alt and title are insufficient to embody the visual qualities of an image. As Clark [1] states, "A longdesc is a long description of an image...The aim is to use any length of description necessary to impart the details of the graphic. It would not be remiss to hope that a long description conjures an image - the image - in the mind's eye, an analogy that holds true even for the totally blind."

Although the alt attribute is mandatory for accessibility and for valid (X)HTML, not all images need alternative text, long descriptions, or titles. In many cases, you are better off just going with your gut instinct -- if it's not necessary to include it, and if you don't have a strong urge to do it, don't add that longdesc. However, if it's necessary for the whole page to work, then you have to add the alt text (or title or longdesc).

What's necessary and what's not depends a lot on the function of your image and its context on the page. The same image may require alt text (or title or longdesc) in one spot, but not in another. If an image provides absolutely no content or functional information alt="" or background CSS images may be appropriate to use. But if the image provides content or adds functional information an alt would be required and maybe even a long description would be in order. In many cases this type of thing is a judgement call.

[1] http://joeclark.org/book/sashay/serialization/Chapter06.html

Related References:
http://www.d.umn.edu/goto/accessibility#alt

[jordanmcclements]

Interesting thread..

I will have to have a look at using longdesc, and disallowing robots from *some* image folders...

[gsharpe]

I believe that Riya Inc. is getting around to solving this very issue. They are already using Likeness searching technology. The ability to harness visual computing via a search of like attributes on images is long over due.

I don't see how they do it but they actually find color variants in an image and allow you to change those to suit your needs and refine your search on texture, color, shade, etc. I think this technology still probably requires some sort of image tagging or indexing, weather its done by a human or a computer it still needs to store attributes somewhere. Heck, maybe it does it on the fly?

Does anyone else know anymore about the technology they're using?

Check it out at www.like.com.

[jtracking]

I wonder what would happen if you placed images you didn't want indexed into the robots.txt file? Never tried it, but just curious.

I've done it, I believe it works. someone's going to prove me wrong, watch!

[jtracking]

Darren create a second folder with the images you want to have displayed in Google, and disallow the indexing of your existing images folder in your robots.txt

perfect WN. Prefect.

[whipnet]

I only have to worry about this on one site, but for us it's very simple. We create a directory for the images we do NOT want indexed and put that in the robots.txt.

There are some images we do want indexed and we placed those in a directory that does get indexed.

This has worked perfectly for years. I have never once seen one of our non-indexed images on a search engine.

Added: I just noticed this has already been stated.

*

[Penman]

It definitely works to use your robot.txt file to keep the image bot from indexing your images. Personally I do not understand why you would care or want your images there. We use a dis-allow all for our image folder in our robots.txt file. We take a lot of our own pictures for our products and our competitors were simply searching them out in Google image and then using them, so we do not allow any indexing of our images now.

If you really want to help your placement try reading this link : http://www.google.com/support/webmasters/bin/answer.py?answer=48367&hl=en
it is information on setting up advanced image labeling for your images through the webmaster tools section so Google better can understand them.

[dharrison]

Hi Darren

I don't think Google can do what your asking, but there is allegedly a new bot for Google images imaginitively called ImageBot. If you read here. (http://www.sitepronews.com/archives/2006/july/19.html)
There is a section that might be of use:


The Imagebot prowls the Web for images to place in Google's image search. Images are ranked based upon their filename, surrounding text, alt text, and page title.

It might not help but it may be a step in the right direction.

HTH.

[kgun]

Disallow:
User-agent: Googlebot-Image

Kgun, if we use that, then all images will be excluded. I think you missed the thread starters question.

Yes, John.


Hi Darren

I don't think Google can do what your asking, but there is allegedly a new bot for Google images imaginitively called ImageBot. If you read here. (http://www.sitepronews.com/archives/2006/july/19.html)
There is a section that might be of use:


The Imagebot prowls the Web for images to place in Google's image search. Images are ranked based upon their filename, surrounding text, alt text, and page title.

It might not help but it may be a step in the right direction.

HTH.

Yes, Debbie that was the bot I mentioned. Good article.

"On the other hand, some web sites may not benefit from Google image search. In most cases, the traffic from the Image search engine is very low quality and rarely converts into buyers. Many people are often just looking for images that they can swipe. So, if you want to save some bandwidth, use your robots.txt file to block ImageBot from accessing your image directory".

Note: ImageBot == useragent: GoogleBot-Image

Important link:
A Standard for Robot Exclusion (http://www.robotstxt.org/wc/norobots.html)

Related link:
Using robots.txt when Multihoming Web Crossing (http://faq.webcrossing.com/webx?eFAQprintFAQ@@.68629d79)

Interesting article:
Observed Web Robot Behavior on Decaying Web Subsites (http://www.dlib.org/dlib/february06/smith/02smith.html)

[Webnauts]

Interesting thread..

I will have to have a look at using longdesc, and disallowing robots from *some* image folders...
Jordan you picked up the cream of the thread. :)

[darren13]

Good info, thanks to all - will look at moving images I don't want people to Goggle at(bad pun, I know) to a separate folder then using nofollow in robots.txt.

Many T's again,
D

[whipnet]

This is all you need in the robots.txt file

User-agent: *
Disallow: /noindeximagesfolder/


The folder name being where ever you decide to place them off your root.

*

[kgun]

Yes and that should block all good (well behaved) bots from the image folder. But there are other image bots like:

User-agent: Road Runner: ImageScape Robot (lim@cs.leidenuniv.nl)
User-agent: Image\ Stripper
User-agent: Image\ Sucker
User-agent: Web\ Image\ Collector

Some of these bots may be bad, but they are not Google related. Bad bots do not respect the robots.txt file and you have to block them in .htaccess.

[whipnet]

Yes and that should block all good (well behaved) bots from the image folder. But there are other image bots like:

User-agent: Road Runner: ImageScape Robot (lim@cs.leidenuniv.nl)
User-agent: Image\ Stripper
User-agent: Image\ Sucker
User-agent: Web\ Image\ Collector

Some of these bots may be bad, but they are not Google related. Bad bots do not respect the robots.txt file and you have to block them in .htaccess.

Yes, you are correct. Thanks for elaborating.

*

[eternity]

Good post by Webnaut on "Images and alt attributes". explains well. his post should be a sticky?

[blitzen]

Are you kidding? You can't control anything Google does!

[Webnauts]

Are you kidding? You can't control anything Google does!
Are you trying to say that Google does not obey to the robots.txt?