SEO Optimization images is becoming more and more essential in SEO (Search Engine Optimization) for websites. The ALT attribute is really a critical step that is often overlooked. This is often a lost opportunity for better rankings.
In Google's webmaster guidelines, they advise the use of alternative text for that images in your web site:
Images:. Use the alt attribute to provide descriptive text. In addition, we recommend utilizing a human-readable caption and descriptive text around the image.
Why would they ask us to do that? The answer is simple, really; search engines have a similar problem as blind users. They cannot begin to see the images.
Many webmasters and inexperienced or unethical SEOs abuse using this attribute, attempting to stuff it with keywords, hoping to achieve a particular keyword density, which isn't as relevant for rankings now since it was previously.
On the contrary, high keyword density can, on some search engines, trigger spam filters, which may create a penalty for the site's ranking. Even without such a penalty, your site's rankings won't benefit from this tactic.
This method also puts persons who use screen readers at a greater disadvantage. Screen readers are software-based tools that really read aloud the items in what's shown on the screen. In browsing the net, the alt features of images are read aloud too.
Imagine listening to a paragraph of text which is then repetitions of many keywords. The page would be not even close to accessible, and, to put it mildly, will be found quite annoying.
What is an Alt attribute?
An ALT attribute should not be used as a description or a label to have an image, though many people utilize it for the reason that fashion. Although it might seem natural to assume that alternate text is really a label or a description, it's not!
What used within an image's alt attribute should be its text equivalent and convey exactly the same information or serve exactly the same purpose the image would.
The goal is to provide the same functional information that a visual user would see. The alt attribute text should function as a "stand in" when the image is unavailable. Think about this: If you were to replace the image using the text, would most users get the same basic information, and wouldn't it generate the same response?
Some examples:
If a search button is a magnifier or binoculars its alt text should be 'search' or 'find' not 'magnifying glass' or 'binoculars'.
If the image is supposed to convey the literal items in the image, a description is suitable.
If it is designed to convey data, then that data is what is appropriate.
If it is meant to convey using a function, then the function is what should be used.
Some Alt Attribute Guidelines:
Always add alt attributes to images. Alt is mandatory for accessibility and for valid XHTML.
For images that play only a decorative role within the page, make use of an empty alt (i.e. alt="") or perhaps a CSS background image to ensure that reading browsers don't bother users by uttering things like "spacer image".
Keep in mind that it is the function from the image we're trying to convey. For example; any button images shouldn't include the word "button" within the alt text. They ought to emphasize the action performed through the button.
Alt text ought to be determined by context. Exactly the same image in a different context may need drastically different alt text.
Try to flow alt text with the remainder from the text because that's how it is going to be read with adaptive technologies like screen readers. Someone listening to your page should hardly be aware that a graphic image can there be.
Please remember that using an alt attribute for every image is needed to meet the minimum WAI requirements, that are used since the benchmark for accessibility laws in UK and the remainder of Europe. They are also necessary to meet "Section 508" accessibility requirements in the US.
It is useful to categorize non-text content into three levels:
Eye-Candy
Mood-Setting
Content and Function
I. Eye-Candy
Eye-Candy are things that serve no purpose apart from to make a site visually appealing/attractive and (in many cases) fulfill the marketing departments. There is no content value (though there might be value to a sighted user).
Never alt-ify eye-candy unless there's something there that will enhance the usability from the site for somebody using a non-visual user agent. Make use of a null alt attribute or background images in CSS for eye-candy.
II. Mood-Setting
This is actually the middle layer of graphics which might serve to set the atmosphere or set the stage so to speak. These graphics are not direct content and may not be considered essential, but they're essential in that they help frame what is going on.
Try to alt-ify the 2nd group as is sensible and it is relevant. There may be instances when doing so may be annoying or detrimental to other users. Then try to avoid it.
For example; Alt text that is just like adjacent text is unnecessary, and an irritant to screen reader users. I suggest alt="" or background CSS images in such cases. But sometimes, it's vital that you get this content inside for those users.
Usually it depends on context. Exactly the same image in a different context may require drastically different alt text. Obviously, content ought to always be fully available. The way you use this example is really a judgment call.
III. Content and Function
This is where the look may be the actual content. Always alt-ify content and functional images. Title and long description attributes can also be so as.
The reason many authors can't figure out why their alt text isn't working is they don't know why the images exist. You have to determined precisely what function an image serves. Consider what it is about the image that's important to the page's intended audience.
Every graphic has a reason for standing on that page: since it either enhances the theme/ mood/ atmosphere or it is critical to what are the page is attempting to describe. Knowing what the look is perfect for makes alt text simpler to write. And practice writing them definitely helps.
A method to check the usefulness of alternative text is to imagine reading the page on the phone to someone. What would you say when encountering a particular image to make the page understandable towards the listener?
Aside from the alt attribute you have a couple more tools at your disposal for images.
First, in level of descriptiveness title is within between alt and longdesc. It adds useful information and can add flavor. The title attribute is optionally rendered through the user agent. Remember they're invisible and not shown like a "tooltip" when focus is received through the keyboard. (So much for device independence). So make use of the title attribute only for advisory information.
Second, the longdesc attribute points towards the URL of a complete description of an image. When the information found in an image is important to the concept of the page (i.e. some important content will be lost when the image was removed), a longer description compared to "alt" attribute can reasonably display should be used. It can offer rich, expressive documentation of a visual image.
It ought to 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 the image...The aim is to use any period of description necessary to impart the facts from the graphic.
It wouldn't be remiss to hope that the long description conjures a picture - the image - in the mind's eye, an analogy that holds true even for that totally blind."
Even though alt attribute is mandatory for web accessibility as well as for valid (X)HTML, not all images need alternative text, long descriptions, or titles.
In many cases, you are best just choosing your gut instinct -- if it's not essential to include it, and if you don't have a strong urge to do it, don't include 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 it is context on the page.
Exactly the same image may require alt text (or title or longdesc) in one spot, although not in another. If an image provides simply no content or functional information alt="" or background CSS images might be appropriate to make use of. However, if the image provides content or adds functional information an alt will be required and perhaps a long description would be in order. In many cases this type of thing is a judgement call.
Listed below are key steps in optimizing images:
Select a logical file name that reinforces the keywords. You can use hyphens in the file name to isolate the keyword, but avoid to exceeding two hyphens. Stay away from underscores like a word separator, like for example "brilliant-diamonds.jpg";
Label the file extension. For example, if the image internet search engine sees a ".jpg" (JPEG) file extension, it's likely to assume that the file is really a photo, and if it sees a ".gif" (GIF) file extension, it's going to assume that it is graphic;
Make sure that the written text at the image that's highly relevant to that image.
Again, do not lose an excellent chance to help your website with your images searching engines. Use these steps to rank better on all the engines and drive more traffic to your site TODAY.
Yankees lefty Pedro Feliciano will get a second opinion on his injured shoulder, but an MRI indicates he's likely done this year.
Peter Risdon Says:
May 16, 2011 at 2:40 pm | Reply
The paragraph you quoted from ends with this sentence:
“According to the best-sited stations, the diurnal temperature range in the lower 48 states has no century-scale trend.”
That was a surprise given the tenor of this post: “… maybe this is the end to questions as to whether surface temperature increases actually exist.”
Did you mean that we can now say the answer to that is that surface temperature increases do not exist? Or that, pace Keenan in the WSJ, the data do not contain statistically significant trends?
andyrussell Says:
May 16, 2011 at 2:58 pm | Reply
I don’t think diurnal temperature range is very important. Do you?
What’s more, the “century-scale” bit covers some interesting detail. Before Fall et al., it seems that the only work on diurnal temperature range showed a negative trend from the mid-century to 1980s-ish. What Fall et al. found was that this has increased again since the 1980s. So there’s no “century-scale trend”.
But that tells you very little about mean surface temperature trends.
Mark Says:
May 17, 2011 at 10:45 pm
I have heard it claimed that the reduction in diurnal temperature range over the past few decades provides evidence that GHG increases are responsible for the warming. In that sense, some people think diurnal temperature range is important.
Incidentally, I don’t think Fall et al. were the first to find that DTR has increased since the 1980s. I read a paper that said much the same thing a few years ago.
Sorry for the lack of references to back up these statements. I’m a little too busy at the moment to chase them up.
andyrussell Says:
May 18, 2011 at 8:43 am
Ok, so I’m probably not giving DTR as much significance as it deserves.
My point is that I don’t really care about DTR. I don’t think I know anyone who has a particular interest in DTR. If this paper had been published by anyone else I wouldn’t have looked at it. It’s not very interesting. It’s just another paper on climate observations that fits in with the “consensus view of climate change” or however you want to put it. That’s useful, but not to me or most people.
If, however, the paper had shown what Watts has been saying it would show for quite a while now (i.e. that the postitive temperature trend in the surface station record in the US was an artefact of poor station siting) then that would have been very interesting. To me and to many other people.
But it didn’t.
Ben Says:
May 16, 2011 at 4:49 pm | Reply
So Peter… If the diurnal high and the diurnal low both rise by 1°C, you think this means there has been no warming? After-all, the diurnal range hasn’t changed! Others might draw a different conclusion.
Peter Risdon Says:
May 17, 2011 at 8:26 am | Reply
I understand diurnal range has significance, and the relationship between day and night time temperature ranges is important, especially with regard to the period 1950 to 1980 when the effect of man-made global warming, it has been argued, was masked by a cooling but revealed by the changes in the relationship between these ranges.
I further understand that this argument is based on the idea that human pollution caused this daytime cooling, that it affected the range of day time temperatures as well as the difference between night time temperatures which continued to show warming, and daytime ones that didn’t. This makes day time temperature range significant: if this is right it would be expected to show a variation that correlates with human activity.
But this isn’t my field; I’m just reading what I can in an attempt to understand as much as possible about an important issue and, for me at least, that means reading Watts and reading this blog. Just searching out stuff you’re already disposed to accept isn’t good enough. My comment was prompted by what struck me as a somewhat partial quotation and exasperation: I’m with Feynman when he said you should point out the problems with a theory, not just the things that support it.
[It's not really a "partial quotation" is it? That sentence you are interested in is stuck on the end of the abstract as a new paragraph and isn't really related to the 2 sentences I quote and which are related to the subject of this post. I'm not really interested in DTR and I doubt Watts was either. - AR]
At least Watts invites people with different views to post on his blog and has been at the forefront of attempts to cross the ideological divide, not least with Judith Curry.
Ben, of course you’re right. Andy, a century is an arbitrary scale, of course.
I’d still be interested in your take on statistical significance.
JMurphy Says:
May 17, 2011 at 12:05 pm | Reply
In what way has Watts atempted to cross “the ideological divide” ?
Ben Says:
May 17, 2011 at 3:16 pm | Reply
Peter, I encourage a critical (i.e. thoughtful) reading of Anthony’s blog but my god do you really think he’s “at the forefront of attempts to cross the ideological divide”? Anthony has done more to harden denialist thought than anyone, with the possible exception Marc Morano.
The “different views” he solicits are unthreatening fig-leaves.
andyrussell Says:
May 17, 2011 at 3:24 pm | Reply
I’ve got no problem with most of what Keenan says, although he’s not the first/only person to be saying these things. There’s a JoC paper from 2010 and it was one of the useful points to come out of the UEA email enquiries (i.e. working more with stats people). Not sure where the funding was supposed to come from for these new people though!
I suppose the bigger problem comes down to climate science covering so much stuff – you can’t just look at problems from a stats/dynamics/modelling/chemistry/radiation/whatever perspective for too long before a) not getting very far or b) needing to doing something you’ve not done before.
omnologos Says:
May 19, 2011 at 7:21 am | Reply
Am surprised nobody claimed it was irrelevant as the US only covers 2% of the globe…
andyrussell Says:
May 19, 2011 at 7:43 am | Reply
Maybe that’ll be Watts’ next move: if you can’t find anything in the US, let’s give Europe a go.
I can’t imagine there’d be much enthusiasm left for such a project, though.