naming images

[Navarone]

I have read that naming your images with keyword names is benefical. For example instead of image1.jpg it might be webdevelopment.jpg. If I took every image on my home page and called them webdevelopment1, 2, 3 etc... (not that I really would) could this be considered spamming?

[ihelpyou]

It could be, although I really doubt the spiders put much priority in the name of an image.

[Spider Man]

I said something similar to this in the google forum so I figure I'll just clarify:

I agree with the above - it's highly unlikely that any major search engine places any weighting at all on image names. If they did then overuse would be the same as keyword spamming and everyone would be creating 1x1 transparent gifs with keyword names and putting them in their pages 600 times (Vive La SSI - dynamically created images) :-)

With google image search the image file name is almost definitely one of the factors that it ranks on (there are several others). In this case changing them will help but the number of visitors you'll get from it is *minimal*. It depends how much you want and need to scrape the barrel!

And keep in mind that in both cases the context which it is in is MORE important than the name itself.

[Navarone]

Well I thought I would ask anyway. Since I am on the subject of techniques, I have also read that heading tags <h1></h1> are important as well. What say ye?

[ihelpyou]

Yes. There are given a extra boost as long as used in a good way. I prefer the <h2> tag myself.

[Advisor]

I think with the naming of images, it's hard to say for sure. It certainly wouldn't hurt you, and who knows, maybe it *would* help. Of course, only if you're doing all the other, more traditional SEO stuff. Image names alone would certainly never get you a good ranking, but who knows, given everything else being equal, maybe that image name could push your site up a notch. Like I said, couldn't hurt. I can't imagine any search engine removing your site because you named your images with similar names. That would be unfair to all those who had no clue of what they were doing, but just named their images similar things. Doesn't FP or some other editor name the images things like image1 image2, etc.?

Hmmm...maybe we should do a search for image2 and see if pages with images named that way rank high. That might be a good test!

Jill

[Advisor]

Well, that was weird. I went to Google and did a search for image2 and got results that didn't appear to use that "word" anywhere, not even in the image names. I wonder how Google decided to rank them under that term. Very strange. There were some later results (mostly 2nd page) that actually had image2 in the title...weird that those weren't ranked first.

And when I tried image1 there was a site that is image1.com that came up number one. Tells you that domain names are definitely important in Google, I guess.

Jill

[Advisor]

Well, it looks like AltaVista probably DOES read the image file names. I found evidence when searching for image2 that they were picking up sites that had image2 as part of their file name. How interesting!

Jill

[ihelpyou]

Yes. Very!

[Advisor]

Interestingly enough, both Excite and Lycos seemed to have more relevant results for image2 than Google. There were a number of pages that used it in the Title, which they picked up, but Google didn't. I didn't check to see if the pages were still good or not, however. But on the surface, they did look to be more relevant.

Oh yeah, and there was no evidence of either Excite or Lycos picking up the actual image file name that I could tell.

Jill

[Spider Man]

Duh-Huh! So why didn't i think of doing that. Must've had my stupid head on that day. As AV etc are covered above - i went back to google and assumed that maybe it was taking "image2" to also be "image" and therefore making it hard to interpret results...

So I found a nice long number that was an image file name on some sites and which google couldn't possibly extrapolate other words from. 'twas 174911.gif. Chop it down to 174911 and if you take a look at most of the top ones and the end ones you can pretty much see that none of them are getting anything from the image file name. Mainly its from body text - otherwise from anchor filenames and text and alt tags.

To double check. I picked a recently listed page of mine. And had google search only that domain for "5579.gif", "5579", "horus.gif" - all of which returned nothing.

In conclusion then, even if you called a file "ilovegoogleohsomuch.jpg" it still wouldn't help even a smidgeon with google rankings.

The jury's still out on what would happen if you called it "ilovegoogleveryveryverymuch.jpg" though :-)

[JuniorHarris]

Curious in Jills' case, if the search results could have been influenced by a link to those domains (results) using an image2 image for the link?

[Spider Man]

I don't think you could ever tell. Would be too complicated. But as we know that alt tags and text of an anchor do count then something like this is quite possible.

<a href="http://www.sitename.com"><img src="image2.jpg" alt="image2.jpg (6345 bytes)></a>

I just think it would take someone a very long time to prove it one way or another. In fact, i'd really like to know this one if anyone can work out how to tell!

[JuniorHarris]

Yes, it could be a very time consuming exercise...maybe something to do on a rainy day.

[Advisor]

JH, that might actually be the only thing that makes sense. I couldn't figure out why in the world those pages would rank high with absolutely no reference to image2 on the page or in the code. And I have seen Google do stuff like that before. I am surprised though that they would give it that much weight, even over a Title tag. In NorthernLight, there were lots and lots of pages full of sites that used image2 in the title tag. Yet these were not found in Google (at least not in the first few pages of results). I wonder what this might tell us about Google and the Title tag. I think S&R would be interested in the conversation! They love this kinda stuff!

Jill

[JuniorHarris]

<lol> I was thinking the same thing...sounds like a job for S&R!~

[highman]

From Googles cached page, ranked # 1 for image2

'These terms only appear in links pointing to this page: image2 '

Investigating the backward links to the page it does indeed use a link wrapped around an image using a mouseover event, the image name is 'image2'

mmm.... it backs up what we already know about the importance of the anchor text in links to your site, but it does add wieght to the theory of getting links into your site named correctly, and that includes the image name of a inbound link.

interesting

[Spider Man]

Taking a quick peeky. they all come from another of their domains and the alt text link in the page i looked at said:

<img name="image2" src="../images/tab2.gif" alt="image2"...

No other text - Thus it's either getting it from the name= bit or the alt tag.

Must look more later...

[Advisor]

How do you investigate the backward links to the site? I couldn't figure that out!

Jill

[highman]

Jill, use you freshly installed google bar, the page info link on the bar provides you with instant 'backward links' info

It also allows you to search a specific site for keywords which is pretty neat

These G o o g l e guys are clever

[Spider Man]

Okay...i tested a few more of the linking sites and couldn't decide anything conclusively (because image2 was also mentioned elsewhere in a lot of the linking packages). So on a different tactic TheCounter.com's code that you put in a page includes the line:

<a HREF="http://www.TheCounter.com" TARGET="_top"><img SRC="http://c1.thecounter.com/id=699885" ALT="TC" BORDER="0" width="88" height="31"></a>

So keeping in mind there's over half a million users there - if it's the alt text of the image in the anchor then thecounter.com should rank really high for the search term TC and with a bit of luck it would be from only linking sites.

And low and behold it does.

I haven't got my logical head on today but i think that shows that the alt text of an image linked to you is a major factor that matters (if not the only one)? Presumably it's the same when linking internally within a site?

[Advisor]

Cool...I had just found that backwards link thingee! Now, is there anyway of excluding the domain of the site you're looking at backward links from? That is, if I'm checking to see who is linking to a particular page on rankwrite.com can I exclude any links from rankwrite.com itself? I remember seeing that you could do this with some engine, but is it possible with Google? If so, how does one do it?

It seems that as part of the PageRank formula, Google counts links from the site itself. So like my inner links page, has a page rank of 7/10 even though no one links to it. Simply because the rest of my site has links to it. I can understand that this would be to stop stand along gateway pages from having a good page rank, but it doesn't seem like links from your own site should count for very much, in my opinion. Now I see why I shoulda listened more closely when people tried to explain all that term vector gobbledy-gook to me. Instead, I pull a Homer Simpson and think of donuts!

Jill

[Sharon & Roy]

Originally posted by webwhiz

Now, is there anyway of excluding the domain of the site you're looking at backward links from?

Hi Jill,

The syntax to exclude your domain is ...

-site:rankwrite.com

The problem is, that it doesn't currently work with the links syntax of ...

link:rankwrite.com

Hope Google incorporates this soon, cuz we sure need it too!

Originally posted by webwhiz

Well, that was weird. I went to Google and did a search for image2 and got results that didn't appear to use that "word" anywhere, not even in the image names. I wonder how Google decided to rank them under that term. Very strange. There were some later results (mostly 2nd page) that actually had image2 in the title...weird that those weren't ranked first.

Jill, what you are experiencing is the fact that Google has a separate database for image searches ...

http://images.google.com/

What this means is that Google doesn't index image file names (i.e. <img src="image2.jpg">) in it's main database, therefore a search for image2.jpg will never return the location of an actual image but just the term image2.jpg as the written word in any other indexed HTML tag.

---

Additional points to keep in mind.

Keep in mind that Google DOES index the Alt Tag Text.

So a search for ... image2 ... WILL return a page with this code ...

<img src="image2.jpg" alt="image2.jpg"> or this
<img src="image2.jpg" alt="image2.gif"> or this
<img src="image2.jpg" alt="this is image2">

... but will NOT return a page with this code ...

<img src="image2.jpg" alt="image3.jpg"> or this
<img src="image2.jpg" alt="image3.gif"> or this
<img src="image2.jpg" alt="this is image3">

---

With that in mind, your chosen search term of ... image2 ... doesn't leave very many pages that were specifically optimized (if any ever were) for that keyword.

Therefore your results will leave much to be desired if you attempt to compare them with other search engines, since that would now be like comparing apples with oranges.

What your chosen search term does reveal though is the Power of PageRank and a further glimpse into Google's current algorithm and where some of it's current "weight" may be placed.

We did a specific search for ... image2 ... in the Title Tag Only which means that only pages with that term in the Title Tag would be returned.

Well as you can see from the results below, there were 8 pages that ranked higher than the first such page.

What that means is that PageRank coupled with the Off The Page Ranking Criteria of Anchor Tag Text or Alt Tag Text currently have enough "weight" to override the Title Tag Text of a page with a PageRank of 6.

This is evidenced by the fact that the top 5 ranked pages don't even have the search term on their page, but it is in the anchor tag text or in the alt tag text of the page linking to it.

See below for additional discoveries.

---

With a search for ... allintitle: image2 (http://www.google.com/search?num=30&hl=en&imgsafe=off&safe=off&q=allintitle%3A+image2) ...

The first page in the results with image2 in the "Title Tag" is ranked #9 a PageRank of 6 - term only found in Title Tag

The second page in the results with image2 in the "Title Tag" is ranked #11 a PageRank of 3 - term only found in Title Tag

The third page in the results with image2 in the "Title Tag" is ranked #14 a PageRank of 3 - term found in Title Tag and as the only Plain Body Text --> image2.jpg

---

Additional research shows ...

#1 - PageRank 7 - not used on page or in URL - Cache says --> These terms only appear in links pointing to this page: 25 inbound links all from the same domain and all linked pages have PR7 to PR5 and <img name="image2" src="../images/tab2.gif" alt="image2" ...

#2 - PageRank 2 - not used on page or in URL - Cache says --> These terms only appear in links pointing to this page: same domain as above, similar case as above

#3 - PageRank 4 - not used on page or in URL - Cache says --> These terms only appear in links pointing to this page: 4 inbound links with the #1 linked page with a PR5 with an anchor tag text of <font size="-1">image2

#4 - PageRank 3 - not used on page or in URL - Cache says --> These terms only appear in links pointing to this page: 0 inbound links

#5 - PageRank 6 - a frameset page; not used on page or in URL - Cache says --> These terms only appear in links pointing to this page: 30 inbound links with #1 linked page with a PR7 and listed in the directory. - #3 linked page with a PR7

#6 - PageRank 4 - used once as Plain Body Text --> Image2

#7 - PageRank 5 - used once in URL --> applepiesolutions.com/image2.wbmp/

#8 - PageRank 7 - used once in Alt Tag --> Alt="image2"

---