Now we are onto the visible bit, you may want to see how a search engine views the body of your website. The simplest way is to use a browser add-on like the one from http://chrispederick.com/work/web-developer/ that allows you to turn off images, styles etc. It’s available for the Chrome and Firefox browsers.
Remember to include text in your pages
Sounds obvious, but there are websites that fail to include a significant amount of plain text in one or more of their pages.
This is particularly catastrophic if one of those pages is the home page. The absence of text makes it nigh on impossible for a search engine to work out what the page and/or site is all about. It can happen if large images, Adobe Flash or Microsoft Silverlight is used instead of text.
Parts of your website may look as though they are in plain text, but in fact that text may be contained within an image, so would be missed by a search engine. Designers sometimes did this to get the font face they preferred, but there are now better ways of applying more unusual fonts to text using web fonts.
Don’t muddle up multiple topics on a single page
This is both an SEO and UXO issue but may be slightly different for each.
Adhere to the principle of separation of concerns as far as possible. You get the benefit of clarity for the visitor and search engine alike.
There is a balance to be struck between presenting each topic singly and having a huge quantity of pages. Visitors lose patience if they have to click too much.
Get a third party to check your content, not least for spelling and grammar
When you are immersed in your business (and website) it’s too easy to make assumptions, drop into jargon or make failed attempts at humour.
A proofreader can ensure that your site is comprehensible to ordinary visitors. Spelling and grammar can be incredibly divisive, so it’s not a bad idea to arm yourself with Fowler’s Modern English Usage to look up accepted forms of English. You don’t have to agree with it, but at least you know where others are coming from.
Good text and grammar make it easier for a machine translation service to convert
The major search engines have translation facilities and clearly it’s going to be easier for them to translate a page with correct spelling and grammar.
Translation technology has to extract semantic information from a web page. This also improves its understanding of the page in the ‘home’ language and therefore the given search engine’s index.
Ensure that your pages contain a good density of words relating to the page topic
It’s easy for irrelevant words to get to the top of the word count. You don’t want to have read more or click here as your most significant words. Click here is also less meaningful for those with mouse-free tablet devices like the iPad.
A browser add-on like Seoquake will show you the frequency of words in your page. In practice it probably means that relevant words and phrases should be repeated four times or so. But this should be within the context of making it readable.
Break up your message into digestible sections
The use of bold and italic (with the
<em> tags) will be picked up by the search engines and will contribute to the hierarchy of the text. Equally, visitors scanning the page quickly are more likely to read a concise message if presented in bold or italic.
So there is merit in breaking your message into at least two paragraphs:
- the first for the person scanning quickly (bold or italic);
- the second for someone who wants more detailed information, with the side benefit of getting your key word/phrase density up – all this extra text is grist to the mill for the search engines.
White papers, case studies, track records and testimonials are a good way of increasing the number of key words and phrases
These things can legitimately hit the spot with both SEO and UXO.
Articles and features do take time to write but they can dramatically increase your word density. As opposed to normal quick-scan, shallow-reading web browsing behaviour, a visitor navigating to a white paper, case study etc will expect to do some reading.
Never duplicate pages intentionally – though it can happen inadvertently
Pages should always have content at least 20% different from any other page on your website and indeed the web as a whole.
Search engines take a dim view of duplicate content and will adjust your ranking accordingly. They also don’t like different domains with the same content or doorway websites of a single page that just point to the ‘main’ site.
One particular complication arises if you use both a global (.com) and a national (.co.uk, .fr, .de) top-level domain name for a single site. This might be because you want to prevent cybersquatting – see the FAQs chapter for more information on this topic.
If you have multiple domains, you need to select a principal domain and then use a mechanism such as a 301 Permanent Redirect on the server to point secondary domains at the principal domain. You may need to speak to your hosting provider or website designer to set this up.
Use the ‘canonical’ meta tag to ensure that your home page is not inadvertently duplicated
Your home page file name will typically be index.htm, index.php, default.aspx or some other. Your web server will be configured to go to this page when someone puts in your web address without any ending.
This means that we have a situation where:
are the same page and therefore duplicate pages in the eyes of search engines.
This naughtiness can be overcome with a meta tag in the header block of the page:
<meta name="canonical" href="http://www.replace-this.co.uk/" />
In addition, your XML sitemap (mentioned in page on links) should not include an entry for the default file name (ie the aforementioned index.htm, index.php, default.aspx or the name specified by your hosting service).
It can get a bit more complicated on some pages. You may have noticed on websites you visit that some page addresses are followed by a question mark and lots of apparent gobbledegook:
The text after the question mark is called the query string and parameters within it are separated by ampersands. Query strings can also lead to inadvertent page duplication as the website theme in the example above is not part of the product. If this website had two user-selectable themes ‘modern’ and ‘classic’ that did not affect the text in the page, then
would have identical text to the modern theme and would therefore be a duplicate page. This can be fixed with a canonical meta tag in the relevant page:
<meta name="canonical" href="http://www.replace-this.co.uk/catalogue/"?id=123&size=M" />
If it is a generated page where you can’t control this and you think there is a problem you will have to contact the vendor of the software you are using.
A favicon is useful for visitors when they bookmark your website
This bit of UXO can be implemented by putting a 16x16 pixel image of your choice in .ico format and called favicon.ico in the root folder of your website.
On some websites you may have noticed a small image, usually a logo, in the browser address bar or on a tab. This is a favicon, also known as a shortcut icon.
Although the ico format originates from Microsoft, it is universally understood in the context of web browser icons. Obviously you will need an image editor that can create ico files and a good one will allow you to put two or more images eg 32px x 32px or 48px x 48px in the same file. This will mean if someone bookmarks your website and places the link on their computer desktop, it won’t look so raggedy.
Icons for the Apple iPhone are larger and start at 57px x 57px. The basic approach is to put a png image called apple-touch-icon.png in the root directory of the website. For more details go to http://bit.ly/d2RIXr to see the advice from Apple.
Not having a favicon will cause a 404 Not Found HTTP status code to be returned as all major browsers now look for them and although the page is rendered correctly, it can cause delays. These become significant on high trafffic sites.
That's not the end of the story on favicons but from a practical point of view, just ensure that you have one and that it's a physically small file.
Always include the year on dates
People may want information within a specific date range – and that may not be just right now – and by including the full date you make it easier for the search engines to classify the page and clearer for visitors.
Websites can have pages that hang around for perfectly good reasons – an archive perhaps – and dates without a year can easily be annoying to visitors. So all your good SEO work leaving the archive around for people to happen upon has been undone.
Aside from that, date formats can be a nightmare because 8/6/2011 will mean different things to different nationalities. So if you have a website for a UK B&B expecting US visitors then they might think that they have booked for 6 August whereas the B&B is expecting them on 8 June. The trick here is always to spell out the month in letters.
With HTML5, it is possible to insert universally-understood dates as meta data, but that’s unlikely to be a good business reason for completely upgrading your website just now. Also, since this is meta data, it doesn’t fix the missing year problem for the visitor so it remains a UXO issue unless remedied.
Make sure your prices are clearly visible
A visitor looking for prices will want to be able to establish the total cost including taxes, shipping to a particular postcode, discounts etc quickly and efficiently. Making this easy for them will encourage them to buy from you rather than someone else.
The visitor should be able to see prices with the minimum of clicks. They may also expect to be able to shortlist, compare items or see a running total as they add (or take away) items in the shopping system.
Avoid putting people off by having long, convoluted paths through the site or making them log in just to get prices. Incidentally, requiring a visitor to register just to get prices can lose about 30% of your prospects.
If you are a wholesaler, make sure the visitor can go through to one of your resellers and that they end up on the correct product page.
Use a print CSS stylesheet to remove items not required when printing your pages
Usually the colours and layout of a website are controlled using a CSS (Cascading Style Sheet) file. You can please your visitors by not emptying their toner or ink cartridges when they print your page by removing solid blocks of colour, the navigation and anything else that has limited use on a bit of paper.
Be consistent if you adopt US English instead of your native version
Some non-US companies adopt US phraseology, spelling and grammar because they think it will give commercial advantage in a global market.
This UXO issue might have downsides in the home market with some visitors. If you choose to go American, then be consistent. Some websites get an identity crisis as a result of mixing native and US English.
PDFs are an extra, not a replacement for web pages
Portable Document Format files have their uses, not least because they have proper pagination, but they are not a good substitute for web pages. The best way to handle PDFs is via a link where it is made abundantly clear to the visitor that’s what they are getting.
Apart from the fact that some people may not have a PDF reader installed – though most do – as a visitor it can be annoying to find oneself unexpectedly in a PDF window.
2020 © Caz Limited