Here at Creare we like to keep abreast of developments in the SEO industry on a daily basis. It’s often said that the SEO industry is one of the fastest moving industries of the tech world, and personally I’ve no doubt that this is true.

The fact that Google has such a dominance in the search world means that all they have to do is make a minor change for countless campaigns to be affected. I saw a story today about how the popular how to site ‘eHow’ had escaped penalties from Google’s various Panda tweaks and updates since last July, and this got me thinking. How long should you leave your SEO campaign to cook for before you review your strategy?

Well, there’s really no right or wrong answer to this and my intention here is to simple get SEO consultants thinking about the issue. The answer largely depends upon how long per day, week or month you spend working on your SEO campaign.

In House

For example, in house executives will probably spend more time on their SEO campaigns, given that they probably form the majority of their job description and so are likely to be reviewing their strategies on a daily basis.

But is this a good thing? While it’s important to take stock of campaigns so that they don’t grow stale, too much tweaking and you could be changing your strategy before it’s even had a chance to yield any results. As fast as Google and the SEO industry changes, it’s important to give your campaigns enough time to bear fruit before declaring them dead and changing tack.

A good example from my experience involved a site which refused to rank for a period of time. I was in two minds about the action to take; whether to continue my current strategy of link building with my existing site or to overhaul the onsite SEO as part of a site wide redesign. Following pressure from my client and a desire to get results for them, the latter was the option I chose, but when I was finally able to come to implementing this I found that the campaign had finally responded to the hard work I’d put in as part of my initial strategy and was now ranking well.

The lesson for me was to manage my client a little better and trust my instincts; you have to be sure that your strategy isn’t working before you rush into changing it.

Agency

So how does this differ to how agency consultants review their campaigns then? Well, agency consultants have the benefit of having the backing of a multi staffed agency with colleagues boasting a wide range of skill sets and specialisms. This, coupled with agencies’ tendency to invest time in research and development means that they’re likely to be at the forefront of SEO updates, changes and developments.

So with up to date, as it happens knownledge of industry changes agency staff are able to make informed decision regarding strategy changes that might be overlooked by in house staff. Basically,by remaining at the forefront of industry developments they’re effectively reviewing campaigns on a daily basis, but only implementing those changes which are going to bring significant benefits.

In conclusion then, I’d suggest that while reviewing your individual strategies is important, and should be done every few weeks or months at the very least dependent upon the type of campaign, just as important is keeping in touch with the SEO industry as it moves on. This will allow you to make surgical reviews of your strategies, instead of looking at everything and wondering what to change for the best.

I read a very interesting article today which, as an SEO consultant, I greeted with an inseparable mix of excitement and dread.

Lance Ulanoff of Mashable fame wrote this article about an interview given by Google’s Amit Singhal in which the search specialist discussed how Google is currently developing a system for a totally new kind of search engine.

Basically engineers over at Google are concerned that the search engine doesn’t ‘understand’ search queries, and that this is leading to inferior SERPs. Put simply, the way Google works is based, as many of us will know, on text and content. Because there are so many websites out there with content matching any given search query, Google has to use other additional ranking signals to really weed out the poorer sites and to rank those worthwhile ones in order of relevance to the search query.

But this method of searching gives no thought to the meaning of any word; it’s nothing more than a very sophisticated content searching algorithm.

To quote a similar example to the one Mr Ulanoff provides, imagine you type ‘what is the tallest mountain on Earth’ into Google. Today, the results generated would include various webpages with content on them concerning that search term.

However, as Singhal explains, Google will simply hope that there’s a page out there (created by someone else) with content on it which matches your search term and answers your question.  In theory, the top Google result might be a page entitled ‘What is the Highest Mountain on Earth’ but which does not contain content which actually says what the highest mountain on Earth is (perhaps it’s a page for use in a quiz, revision questions etc).

This then, is where Google’s new ‘knowledge graph’ comes in. Basically, Google are creating a giant repository of millions and millions of words and their associated meanings, or ‘entities’. Using this knowledge base, Google engineers hope to be able to refine the search engine so that it actually understands the search query and all of its many possible connotations, meanings and metaphors. This would mean that it can provide the precise result that you were looking for.  So in the above example, Google would know that the tallest mountain is Everest and that anyone asking that question will probably be interested in statistics about the size of the mountain and details of its location.

Instead of providing a page for you and hoping the answer is on there somewhere, the search engine would presumably be able to present you with the answer directly. If course this has massive ramifications for SEO, but the most apparent issue is that if this was implemented successfully it would fundamental change what Google actually is. Arguably, this change would bring Google into direct competition with knowledge engines such as Wolfram Alpha.

Of course, your average keyword such as ‘garage doors London’ might remain unaffected, but some of your longtail phrases, such as ‘How To Fit a Garage Door in London’ could be demoted in search in favour of knowledge from Google him/herself.

It’s easy to see why this project is so attractive to Google – the hype surrounding Siri for the iPhone 4S (which is different technology to this) shows how receptive people are to AI in theory. All Google would have to do is slap a likeable face, name and personality on the search engine and they’d have cemented their dominance in search for another two decades.

Imagine a search engine that knew (really understood) your likes and dislikes (personalised search) and could use this history to accurately anticipate what you were looking for using meaning as the human brain would, rather than just words.

Apparently this knowledge graph is already in use for some search terms, making basic connections in meaning I order to determine what images are displayed for certain terms. However Google’s game plan seems to be a much more complex and thorough integration of knowledge based search, which could take years to perfect.

For us in SEO, this could spell the end of optimisation techniques as we know it. Arguably, we could lose a huge amount of traffic, as queries are answered by Google itself rather than our page. This would reduce sales and enquiries from impulsive buyers who land on your site through a longtail article page and then go on to make a purchase.

Knowledge based search could conversely improve PPC’s prospects for the future however. If Google is answering more queries directly, then people will be spending more time on the SERPs themselves, thus increasing the value (and going rate) of paid advertising on these page.

Of course, I’m assuming here that Google will answer queries directly itself, which may not actually be the case. Instead, this knowledge might simply be used to more correctly identify relevant pages and therefore provide better SERPs.

Taking our earlier question as an example (what is the largest mountain on Earth), the new Google might actively search for the words contained in the answer to the question as it would understand the question and know its answer. It might search for words such as Mount Everest and Himalayas and return pages contain data on those.

I could probably go on forever on a subject that could fundamentally change SEO, but I’ve gone on for long enough. It’s simply an interesting concept that’s well worth keeping your eye on if you work in the industry.

In keeping with my habit of keeping you all updated regarding Google’s algorithm changes, I wanted to mention a few updates that took place as recently as over the weekend.

The main gist of the updates seemed to be that freshness remains important, and the aim of achieving more recency and relevance in the search results.

Of the freshness tweak, Google said:

We made several adjustments to the freshness algorithm that we released in November. These are minor updates to make sure we continue to give you the freshest, most relevant results.

So not saying much then. Still, this does show that Google seems to have noticed that their freshness update back in November didn’t quite work as planned.

Users complained that the SERPs were displaying less relevant listings simply because they were more recent than the more relevant content, which might have been a little older. The problem was that the original content for any given issue is inherently going to be the oldest content for that issue, meaning that it isn’t likely to fare very well under Google’s freshness update.

Google did mention that they’ve improved the way in which they ascertain the age of a page, and they point to pages discussing recurring events as beneficiaries of this particular element of the update.

So hopefully this latest tweak will work to counter the problems associated with freshness, although evidently we’ll have to wait and see. Keep an eye on your campaigns as this tweak could lead to the resurgence of content you thought far too old to rank well again.

So what else do we have in this latest box of goodies?

Spellings

Well, Google seem to be particularly concerned about spelling right now, with several changes to the way Google’s autocomplete feature makes corrections for common and rarer misspelled terms.

Panda

They’ve also apparently integrated Panda and everything it stands for more firmly into the Google machine, meaning that, in theory, higher quality sites will be ranking more highly in the SERPs. I haven’t noticed any impacts either positively or negatively on my campaigns yet. I think it’s safe to say that if your sites were hit after the main Panda update and you haven’t rectified the problems yet it might be worth sorting, as it doesn’t look like Google are going to drop Panda any time soon.

Picture Perfect

This change relates to how images are ranked in image search. Apparently, if the landing page on which an image is displayed if off a high standard then the image is more likely to rank well. This means that it’s no longer practical to slap that soon to be viral image on a basic page – you’ll need good quality content accompanying it to rank better.

Realtime & News

The final noteworthy SEO related change is a supposed improvement to news article integration in the SERPs. Google have refined the way their algorithm decides which queries should return news results with a view to getting it right more often. This update is specifically significant as it comes at a time when relations between Google and Twitter is at an all time low – suggesting that re-integration of Twitter’s real time results into Google search isn’t likely for a while yet. It’s probably that Google is doing everything it can to improve the real time aspects of its SERPs in order to make up for this loss.

If you’ve been involved in the SEO industry for very long at all, you’ll undoubtedly be aware of the importance of robots.txt files in certain circumstances.
Robots.txt files can be very useful in determining how your site is viewed by Google and the other search engines out there. Not only can you forbid the search engines from seeing (and therefore indexing) your site at all; you can also hide specific pages from their view.

For bigger sites, such as ecommerce platforms with hundreds and hundreds of pages, this can mean configuring quite a sizeable robots.txt file. And this is exactly where the problem can occur. Via his Google + page, Googler John Mueller recently reminded SEOs that, ideally, robot.txt files should be no larger than 500kb in size.

According to JM, Google’s web spider, or Googlebot for short will only read the first 500kb of any robots file. This means that if you have important information in your file, but it’s larger than 500kb then you could be leading Google to disregard said information. This could lead to serious problems and could even mean that your robots.txt file doesn’t succeed in performing the task that you originally configured it for – potentially causing big problems with your SEO campaign.

More than this, an unnecessarily large robots file could be exactly that; unnecessary and could suggest that the entire build and coding of your site might need a rethink. With Google and SEO in mind, remember to check your robots file after any work is done on the site, and remember that the simpler your site is to navigate, the better chance Googlebot has of doing so successfully.

Google have announced a nice update to Webmaster Tools this week that I’m quite pleased about, and sometimes I wonder if Google follow the bad news, good news rule as this welcomed update comes after a potentially controversial algorithm change which Joel blogged about a few days ago.

Previously the sitemaps section in Webmaster Tools simply allowed you to submit an xml sitemap, and displayed the number of pages submitted alongside the number of pages indexed. That was it basically, and the fact that this feature had little data to get you excited about could misguide someone into thinking that xml sitemaps are not too important when it comes to SEO.

I certainly witnessed the power of sitemaps last week when a designer asked me to look at a new site they had just inherited which didn’t even rank for their own brand name – which was not keywordrich.com by any means. After a quick xml sitemap submission in Webmaster Tools, the site jumped straight up to position 4 on the first page of results within 2 days.

Content-Type Data

What’s really great about the latest version is that there is a dashboard which displays the data separated into content types – web pages, images, video and news. For each content type you can see the number submitted to Google, and the number indexed by Google – and even presents this information in a colourful bar chart – data-obsessed folks love charts right?

Sitemap Errors

Another nice thing about the new feature however, is the data in the table below the graph. Webmaster Tools now show the different sitemaps that you’ve submitted, and the number of errors or warnings for each one, if there are any. It helps you to see the types of issues better apparently, although I haven’t seen this for myself yet (phew!).

Test your Sitemaps

Probably the best thing about the update is the ability to test your sitemaps before you submit them. So you can upload one and test whether it has errors or warnings and then fix them before submitting it to Google. On their blog, Google say it only takes a few seconds for the test to come back with results which sounds good, but they’ve also stated that not all errors or warnings will show up in the test, so you should probably check Webmaster Tools after you’ve actually submitted your sitemap to make sure.

Now with Google’s latest update to the sitemaps feature, I think this is the attention xml sitemaps deserve. Oh, and it’s worth noting that the video sitemaps feature in Webmaster Tools is no longer, now that this data is shown in the main sitemap section.

What feature in Webmaster Tools would you like to see updated next?