Website Audits

Search Engine Ranking takes stock of where your site is currently. We analyse a site on a set of agreed keyword phrases and look at its performance and that of the competition.

The reports are very detailed but straightforward and give you an instant picture of where you are in the rankings.

The reports are best viewed after a few weeks once you have begun to build up some data. This begins to show trends in your ranking.

There are a body of people who begin to become obsessed with ranking checking it on a nearly hourly basis. The Optimisation of a site is, in the main, for the long term. Changes in your ranking on an hourly basis may have something to do with your site but are most likely, unless the change is massive, changes to the rank of other sites which kick you up or down.

Large changes in your site can be because of something you have done on the site. Perhaps you have decided to do something less than ‘ethical’ cloaking, link farm etc. In the case below it was a change on the structure of the algorithm that caused this quite spectacular drop.

This site moved from #8 to #86 in just 10 days. It wasn't anything the site owners had done, it was an update to the algorithm that downgraded a page that ranked for a popular keyword phrase. 

This is the very basics of the search process. Google and other search engines use a 'spider' to venture forth and collect information on pages.

Once a page is found all the links on that page are analysed to find further pages on your site.

After some time this leaves a search engine a list of URL’s related to your site and this should be finite (at that particular time). In order to make sure that the spiders have a complete list of pages you should also submit a sitemap to the search engines so that they have the complete list.

The search engines will then begin to apply formula to your pages in order to allow them to grade and rank your pages. Formulas will also be applied which will determine how often they will visit your site (to get new pages or updated information) and how many pages they will look at on each visit.

All that information is available (after a while) in your Google Webmaster Tools account.

A search engine then analyses each page to see what words and phrases appear on a page and, interestingly, in what position. They also look at key content tags such a titles, H1’s, alts, etc (that’s why they are so important). If you want to see how your pages appear to a spider, you can use a Lynx Viewer. This will give you a good idea of how a spider will see your page.

See blog article of a real life example of the difference between the human and spiders view.

So at this point a search engine will have a lot of data about your pages. It might know, for example, your site is about nuts and bolts and that you have a page about ‘blue bolts’. So, when someone searches for “Blue bolts” a search engine knows that your page should be included in the results, but in what position will your page be?

That’s determined by the search engines. The Google Algorithm uses over 200 different factors to determine the pecking order of your page in a search result. The SEO process helps to determine what those factors are and the importance of each one.

One factor they have never minded admitting was page rank. That is the importance of a page based on the incoming links from other pages. That’s where link building comes into its own. Link building cannot/should not be done until your ‘house is in order’.

Page titles and descriptions are your "elevator pitch" in the search results and you only have an instant to make a good impression on a potential visitor.

Page titles are incredibly important in terms of SEO and the title of a book they inform the potential reader as to the pages contents. Search engines it appears place a good deal of weight in titles and the keywords in them. Simply by manipulating the pages titles can have an immediate and positive effect on your search rank.

Actually seomoz, the Search Specialists conduct an annual survey of around 80 of the SEO heavy hitters worldwide to see what they believe are the most important factors in aiding a site to rank. Last year, of all the “on page factors” possible, keywords in a page title was ranked number one.

Looking at this part of the process is an exercise in futility unless you have both analytics running on your site and have signed up to a Google Webmaster Tools (GWT) account as they are two great aids in looking at a sites titles and descriptions. If either of these is only being set up then, ideally, some time should have passed so that relevant data can be gathered before you scrutinise the data.

Using your analytics account you will be able to see the keyword phrases that brought you traffic (and, with a bit of wizardry, where your site ranked at that time). However, it does not show where your site appeared in a set of search results, but where your site was not chosen – which is a much more interesting prospect. Your site appeared in the results and was overlooked.

Google webmaster tools is excellent at giving you some inspiration here. In their query section you can see the search queries where your site appeared, in what position, how many times and, vitally, what your click through rate was for that expression.  Clicking here will show you in a pop-up the table you can see through GWT query. If you have a site with a lot of impressions or appearances but with a low click through rate, then its down to further investigation.

I’d be inclined to look at keywords that are high volume and of high importance and see where your site is appearing and what’s the CTR is on those keywords. If it’s a poor click through rate do the physical search yourself and make an informed judgement as why other sites, both above and below you, might be taking the traffic instead of you. Don't assume that sites below you are getting less or no traffic.

Search marketing company Enquiro did a survey some time back to see how engaged a searcher was with a set of search results. They concluded that a searcher will spend no more than ten seconds looking at a set of four or five results. That’s just two seconds per result in which time they read your page title, your description and possibly scan your URL. Then they weigh up all that information and very promptly decide if your site is going to fulfil their search intent.  So, to spell this out, you have two seconds to make an impression to a potential visitor and that’s based on the shop window of your page title, description and URL.

You can make great strides in SEO amending the page titles on a site. All page titles have to be different and they have to be keyword rich – but not stuffed. Page descriptions are not used but the search engines in their algorithms to rank, but the end user will rely on them when they are scanning a set of search results.

A quick way to see how your site appears in any search results, without going through each possible key phrase, is to use the ‘site operator’. In the Google search bar, key in Site:mysite.ie (obviously changing the name of the site) this will show you all the pages that Google has indexed and how they will appear in search results.

I was looking at a few sites recently and came across one site that was full of information. But it let itself down badly in the search results as not much effort (aka no effort) had been made with the search titles. Not only are the page titles unappealing to the human visitor, the site wont rank with page titles that have no keywords in them.     

Many of the available CMS systems will, if not instructed otherwise, produce URL’s that are long and very forgetful. Moreover, using the base url’s in a CMS system can sometimes create issues in that it can produce different URL’s all pointing at the same page which in turn can create duplicate content problems.

It is better that URL’s are re-written to something that is ‘digestible’ and more agreeable to both the consumer and the spider.

Using words (keywords) in your URL’s has a positive effect on your SEO efforts. Google suggests (a.k.a. warns) that you shouldn't go overboard on the length of the url or stuff keywords into it.

The url issues are somewhat in the same arena to the site structure and also has a foot in the robots.txt section, so both of those sections are worth a look.

URL links such as these are the default here:

http://www.ilevel.ie/index.php?option=com_content&view=category&layout=blog&id=41&Itemid=230

This a few tweaks, your URL's go to this

http://www.ilevel.ie/blog/print

This is infinately better.

The advice would be to keep the structure simple. Hyphenate words with a ‘-‘ blue-widgets as opposed to blue-widgets. URL’s that are dynamically created need to be watched as they may produce different URL’s for different pages.

If you are using an Apache server many of the URL parameters and issues can be amended in your .htaccess file. But a cautionary note: If you have a site that is doing ok in search the work on the URL’s may be very timely then maybe it’s as good to leave well enough alone. To be frank - spending time turning www.mysite/products/product_detail.php?product_id=1 into www.mysite/products/product1 is wasteful if you set it as a priority.   

You may just get a decent enough bounce by concentrating on the more accessible SEO techniques without having to venture into this area too much. It can be very complex.

Really this should be looked at during the development stage of the site – when it’s first being worked through. If it was looked at and planned for at this point, then some of the nasty looking URL’s may not have to make an appearance.

Page Speed has, of late, become particularly important. You can see this reflected in the many new applications that Google have released in the past year. Their page speed extension for Chrome and the larger part that page speed takes in its Webmaster Tools shows that it is putting greater weight on the speed element of web browsing.

A further and by far more interesting application is in their latest edition to their analytics package where there is now a specific ‘page speed’ section in the software. It tracks the average time it takes to load a sample of your pages over time. You can then cross reference these against certain dimensions such as browser or geographic location to see it there are any marked differences.

Then, as you try and lighten the page loads, this should be reflected in your analytics with the average dipping.

There are days where the data may throw you, ever so slightly. The grab below show that one day, where one site spiked in page average, we found that the average was affected because of one “visit” where the page load time for that visit was 706 seconds or about 12 minutes! This wasn't reflected in our server logs, so we are at the mercy of Google on that one.

While Google provides excellent analysis, it doesn’t give you the why. What exactly is slowing down your page and what can you do.

There is always a latency in your analytics data, so if you want to get an immediate view of any changes you make to your pages you can use some of the online resources.

There are plenty of tools available to analyse pages and give you some answers.

Pingdom Tools is one such site or, if you want a bit more analysis, go to webpagetest.org. The latter gives very detailed analysis of all the elements on a page and their page load times. The nice advantage of Pingdom tools is that it keeps a record of each time you analyse your site so you can look back to track your progress. To combine both and get some good tips for solutions gtmetrix.com is not bad at all.

Don't get hung up on the home page - it's the least of your worries. Many of these sites give you the opportunity to, once the main parts of the pages have been loaded into your browser, to revisit a page. This shows you the elements that will be reloading on every page - were their problems there?

Because there are so many elements on a page and so much to be loaded you need to know which ones are performing, which are not and assess what can be done.

Items that can give you a fairly immediate reduction in clutter would be:

• Look at your hosting - dont simply opt for the free version. 'Free' can be very expensive.

• Look carefully at your images and make sure they are an appropriate size. Don't use HTML to reduce your picture size, have them reduced before.

• Minify your js files. This essentially strips out the whitespace in the files but makes them function the same, but they can be much smaller. You can have a look here at the minifyjs.com site where you can do this on line.

• On a point of caution: Back up theses files first. It only takes a few seconds and will save tears if something goes wrong.

• Like minify, compress your css files to do the same. Even template sites can do with a little tinkering on the css front.

Also with the css files you should (if you are on an apache server) try gzip. You can test to se if your site is compressed using a gzip test and hopefully you can get results like this:

Original Size: 31.79 KB
Gzipped Size: 6.72 KB
Data Savings: 78.86%

So, your efforts are rewarded in this case.

This site works off a template that allows gzip to be turned on or off. Without gzip the site load times (for the front page) are in the region of twelve seconds (the front page is a bit ‘phat’ and its unashamedly design over function on that page). Turning on gzip will knock nearly three seconds off the page.

Original Size: 25.25 KB
Gzipped Size: 6.05 KB
Data Savings: 76.04%

So you can see immediately for the above the benefits of gzip.

Sitemaps are there for the benefit of both our human and not so human visitor.

For the Spiders they are a 'pointer' as to where to find pages on the site. It's very likely that spiders will, through your internal linking find all the pages, but it’s better to cut down on the hit and miss and give the spiders a road map of where your pages are.

Most CMS have plugins or widgets to help you draw up a sitemap. If they don’t there are plenty of places on the net what will do it for you automatically, like xml-sitemaps for example. You simply leave the resultant file on the root of your site and then nudge Google through Webmaster Tools that the map is available.

There are plenty of reasons that this file is fairly important.

Firstly it determines the canonical versions of your site, whether you’ve opted for www or non www.

If you can, try and use a sitemap generator that supplies a ‘last modified’ stamp – the actual time/date that the pages were modified. If that’s in your sitemap, it indicates to Google the pages that should be given a priority as they have been recently created or amended. As Google only assigns a certain amount of time to each site at each pass, it important you nudge the spiders in the right direction.

404 Pages appear when one of the pages from a search result (or otherwise) no longer exists on your site. The server normally throws a very basic 'death notice' at the screen and everything comes to a stop - and that includes the momentum for the search.

To overcome this tragedy you really should make a custom 404 page or have a redirect. 404 Pages are important for two reasons. Spiders hate dead-ends and have to have something to guide them on their merry way, as do the human searchers, the letter being particularly fickle.

If a page is missing, for some reason, it’s tragic just to get a some standard system error and a sure fire way of having people suspend their search. Two ways to overcome this:

Redirect any 404 errors to your home page. That’s a very simple (if a bit rough and ready) solution and can be done through your .htaccess file. But the drawback is that it could confuse the searcher as they are unaware as to why they have landed on this page when they specifically asked for another.

The 'correct' way around that is to create a custom 404. It can be based on your page design or it can be completely ‘radical’. The page should inform the searcher that the page they were looking for is missing and, importantly, the 404 page offers them some alternatives.

We use a very simple 404 page on this site but it seems to do the trick as the bounce rate on the 404 page is very low suggesting that stalled searchers do take the opportunity to refine their decision and look for something close to what they were originally looking for if that opportunity presents itself.

I’d also advocate that you make sure your analytics code is in that 404 page.

If you are looking for some creativity pointers for the 404 pages there are plenty about. Try a Smashing Magazine article on 404 pages.

This is a small file that points and guides spiders as to what they can index and what they should keep away from.

There may be areas on a site that that you want to block off from spiders so that that information in those sections does not (or should not) end up on the web. CMS systems have areas that you don't want found on the Net so you use the robots.txt file to ensure that.

Wordpress, for example, suggest you block about 13 folders through your robots.txt file, Joomla about 10. We have tagging on the blog portion of the site for pages. Because of this you could reach the same page in a number of different ways and thus a number of url's. Therefore we have the labelling URL’s blocked so that they can’t be indexed and the page doesn't end up on the web with two url's and giving us a duplicate content issue.

You have to be careful of the syntax in the file for example:

User-agent: *

Disallow: /

This blocks everything, it instructs robots not to index anything

User-agent: *

Disallow:

Without the trailing slash, this now means index everything. It’s only a subtle change and the use of the word disallow would logically suggest that spiders are blocked, but they are not. You can have a look at the syntax used here.

While the robots.txt file will ask the robots to conform to your request, that data may still end up indexed on the web. The best way to ensure that your data is kept safe is not to have it on the web if at all possible. If it is necessary to have these files on a server you might consider password protecting them through something like .htaccess.

If you do make changes to your file, make sure that it is instructing the spiders to do what you want. The best way to see what they will crawl and what they will not crawl is to run your site through “Link Sleuth”. It follows your robots.txt file and will therefore show you what’s up for grabs and what's not.