Industry News
23 February 2010
Is personal search about semantics?
I was recently reading an interview with a director of Bing, Stefan Weitz about the future of search. Stefan has said that one of the biggest problems that search engines have is trying to ascertain the intent of the searcher. This problem has been around for ages and the example we use at Bruce Clay to properly illustrate it is; if you type in java in the search box, are you trying to find information about the coffee, the programming language or the county? Without any additional keywords within the search query, that intent can be difficult to discover.
Photo by dullhunk via Creative Commons |
The next minute I was reading yet another article on personalised search and had a thought, maybe the predominant reason Google instituted personalised search was to better understand the intent of the searcher based on previous searches in terms of language surrounding the query. For example if your personalised searches show 20 previous queries relating to programming language with no queries relating to coffee or an obscure Indonesian island, then it could be assumed your single "java" query would in all probability be related to the programming language. Stefan argues for the need of the search engines to put context around the language and perhaps Google is using personalised search to do just that.
In fact, in the Google video announcing personalised search, they mention that one of the things they hope personalised search will improve is the ability to better determine the searcher's intent. The point I'm trying to make is that with all the negativity surrounding personalised search, perhaps its benefits, particularly surrounding the better identification of intent has been glossed over.
During the course of a regular day I run alot of queries through Google and although I dont have any concrete evidence, I do feel that since personalised search has been turned on by default, my SERP's have got a little better, especially with the search engine seemingly reading my intent better.
While we are on the topic of personalised search, Rand over at SEOmoz had an interesting idea; essentially, he says, "Personalized search is now on by default. This means that every click, branded search, and expression of a "brand preference" or "brand affinity" in Google's results is likely to result in preferential biasing towards that domain in future searches. A "Google" Pontiac message during this Superbowl wouldn't just send users to their site, it would also mean that tens of millions of searchers would now be "personalized" towards that domain."
It's an interesting idea, if you were sure that your company comes up first for a brand search i.e. Nike, then maybe these big corporations should start having "Google Nike" instead of the URL on their brand messaging. This would lead to more people searching Nike and therefore possibly meaning that the Nike web site would come up tops for the search "running shoes" due to your personalised search history.
There are a few caveats for recommending this:
1. Google can decide that personalised search is actually making things worse as supposed to better and just turn it off
2. Users who are not logged in will only have their history tracked for 180 days so you may need a sustained advertising campaign to take advantage of this
3. If you don't appear for your suggested keywords in the organic listings you risk sending a different site your traffic
4. If you are paying for your clicks (eg "search potential" by AmEx) then the traffic to your site is costing you twice (once for TVC, once for adwords click)
There has been a lot of debate within the SEO industry as to what personalised search means. I think anything that helps match a query with better SERP's, is a good thing and to be honest that's why I think Google turned on personalised search for everyone.
Posted by Marc Elison on 23 February 2010 at 11:09 AM | Comments (0) | TrackBack (0)
See more entries in Industry News
18 February 2010
The State of the Australian Search Market
Australia has always had a healthy obsession for a few things that make this country great. Things like your favourite sports team, a great day at the beach and a backyard BBQ, are topics that can consume a large amount of time and conversation. There is one oddity that continues to boggle my mind, which is the focus on PPC when it comes to online marketing spend with lower regard for other search activities, like SEO for example. Even now with more and more knowledge emerging it seems that PPC still has the advantage on SEO in Australia by a large margin.
When online marketing came to the fore a few years ago, the benefits and how to implement PPC were almost immediately understood while SEO was portrayed as a mystical black art, with a bunch of geniuses sitting in a closeted room deciding the movements of Web sites up and down the rankings. Even now with more and more knowledge emerging it seems that PPC still has the advantage on SEO in Australia by a large margin.
Specific search spend figures for the Australian market are a little hard to come but IAB reports that the total search spend was around $900 million for 2009. Of that figure, the brain's trust at Bruce Clay Australia believes that approximately $100-$135 million is spent on PPC management fees (that doesn't include the actually money spent with Google and other services for actually buying the keywords) while roughly $30-$40 million was spent on SEO billings. SEMPO International 2008 data indicates that out of 100 percent of search spend, 87 percent is spent on paid placement (PPC) while the remaining 11 percent was spent on SEO (the remaining budget was spent on paid inclusions and SEM technologies). Yet PPC is said to yield a much lower volume of traffic. It is estimated that it is that 70 percent of Google's search enquiry click-throughs are to search listings provided by the 'natural' search results. So people are more likely to click through to 'natural listings' and it has a significant long term ROI benefit so why do companies and agencies insist on using the majority of their online spending budget on PPC .
 |
Well for one PPC is a lot easier to explain and understand than SEO, especially when you are trying to explain the differing concepts to a client with limited expertise. Spent $5,000, made $20,000 is a lot easier than explaining that the new optimised content you have written had gained 1,000 visitors and 875 uniques with a decreased bounce rate due to the longtail traffic targeted. Another reason for the promoting of PPC over SEO is that the ROI is actually easier to measure with PPC campaigns. PPC is more like traditional billboards than SEO so traditional advertising agencies have an easier time understanding it as an old fashioned media spend and it's easier to recommend and deliver to their clients.
SEO also needs more feedback and initiative from the clients' side. It means working with the client to conduct keyword research and may even require substantial technical investment on the client's Web site to make it 'SEO friendly' whereas with PPC, clients can have a more peripheral role with the running of the PPC campaign with less demands placed on their time and resources.
Yes, PPC results are immediate and the budget can be controlled. Yes, you can geo-locate advertisements to appear in certain states and towns. Yes, PPC campaigns are relatively easy to implement and fairly easy to measure. However, many reports indicate that the long-term ROI of PPC campaigns are limited, while the opposite occurs with regard to natural or organic traffic (as can be seen from this graph to the right from elliance.com).
The good news is the tide is turning, a quick Google trends search worldwide shows how SEO has slowly eroded the lead that PPC has had and even overtaken it (red is "PPC" while blue is "SEO"):
The chart below is for Australia and shows a similar phenomenon as to what is happening worldwide (red is "PPC" while blue is "SEO"):
A report from Eyeblaster in October 2009 found that "Australia and New Zealand are among the least likely nations to click or dwell on online display advertising". The explanation given by Eyeblaster was as follows "In less mature markets, where users are novice in online advertising, they receive ads with more open arms. When the market matures, the users' fatigue together with increased competition for their attention reduces engagement significantly. The users' propensity to engage with ads should be taken into account when comparing benchmarks between geographies and verticals."
Essentially Australia is becoming more 'ad-blind' and this is extending to PPC advertisements. SEO has traditionally been giving a bad rap, and the constant flaming efforts by some people in the industry to create linkbait certainly have not helped. However, there is a movement growing behind the art of SEO these days and education is the key. More and more CEOs and CMOs are becoming educated due to more and more SEO information filtering into the mainstream news. More SEO companies are also taking it upon themselves to educate potential clients through free and paid training sessions. The decision makers are beginning to see the value of SEO over PPC, if they need some further proof, here is a great infographic from DIYSEO.
PPC is also becoming more expensive. With more and more companies becoming active and visible in the online realm, more bids on terms will push prices for high visible PPC positions higher and higher, good for Google, bad for companies using PPC to advertise their wares. In fact research suggests that the cost of popular keywords on Google has increased by up to 300 percent in the last three or four years in Australia. SEO has been criticised for its lack of transparency in the past, i.e. how and what exactly we do to help our clients achieve SEO key performance indicators. The industry as a whole has recognised this and most reputable SEO services providers have tried to make their workings as clear and transparent as possible to appease any uneasiness associated with the so-called "smoke and mirror" approach of the past.
I have a dream. I have a dream where PPC and SEO can exist peacefully and co-inhabit our online world; where these two technologies will be used together to create a symbiotic relationship that will increase ROI for the companies and agencies using them. A recent report by Frost & Sullivan states that "Search Engine Marketing (SEM) and Search Engine Optimisation (SEO) continue to attract advertiser spend in Australia and in a change to previous patterns, the two strategies are becoming more widely adopted simultaneously rather than exclusively" and I think that's a pretty good direction for the Australian online market to be moving in. Studies shown by iCrossing show that clicks, actions, orders, page views, visitors and time on site all increased when SEO and PPC were used together in a complimentary fashion.
I myself am an organic search evangelist. It's something I live and breathe every day. Sometimes I want to shout from the rooftops Gordon Gecko-esque "SEO is good", and a strange shift is slowly happening. Many other people around Australia are starting to feel the same way. We can see that by the way that SEO is catching up to PPC in not only ad spend but searches and mentions around the Web. The more transparent we make the SEO process and the better the education and news around SEO becomes, the more traction it will get and I can't wait for SEO attain its true place alongside PPC in the Parthenon of online spend.
Posted by Marc Elison on 18 February 2010 at 9:54 AM | Comments (0) | TrackBack (0)
See more entries in Industry News
10 February 2010
SEO Factors and Trends Report Released
Bruce Clay Australia is proud to announce the release of our "SEO Factors and Trends" report. This report is a concise yet comprehensive review of the more important SEO ranking factors and the changes made to these ranking factors over the past 12 months or so.
The most important 12 ranking factors are thoroughly reviewed and implications for changes in these ranking factors are also discussed. Ranking factors deemed not vital but still important to your search engine optimisation strategy are also listed and additionally there is a preview of what 2010 might bring us in terms of the search landscape. Download the pdf version of this report from the SEO Factors and Trends report page.
Posted by Marc Elison on 10 February 2010 at 11:47 AM | Comments (0) | TrackBack (0)
See more entries in Industry News
14 January 2010
Deteriorating search engine results
Photo by _Tophee via Creative Commons |
The other day at the Bruce Clay Australia offices we had a discussion about how the search results seem to have deteriorated over the past 3 - 4 years. Whether it was searching for cheap hotel rates in Melbourne or trying to find a review for a laptop, it seems more time has to be spent and more searches have to be carried out just to find the relevant information. Now there are a few reasons why this might have happened and I am going to explore a few of these.
Heightened expectations
With all the advances in technology and search engines algorithms, are we expecting too much from the search results? In the past few years, everything has been geared around instant gratification (I think this was a driving force in the development of real-time search) and we are not only expecting results accurately but immediately. Now I struggle to recall my searching habits of 4 years ago but I have a sneaky suspicion that I was willing to cut all the big search engines some slack when it came to searching. I don't think they have that luxury any more.
I think a big part of this heightened expectations comes from knowledge. I know Google has 5000 PhD's just sitting around thinking of ways to improve the algorithm, but when I want to get the cheapest accommodation in Melbourne I dont want to have to navigate between 10 different, yet almost identical aggregator sites in the SERP's before giving up due to frustration and ending up in Wogga Wogga, just because it's easier to find a cheap hotel.
Another interesting observation is that more and more users are using longer (longtail) search queries to find what they are looking for. The question I pose is: are users searching with more specific search phrases because they are more savvy to the way the web works, or is it because the search results are so poor for broad-based queries that in order to find something relevant you are forced to use detailed and lengthy search queries?
TMI
Too much information! Normally a teenager's response to a friend giving one personal detail too many, it can also describe the explosion of web sites, blogs, aggregators and web properties in general. Since March 2005, the number of people using the Internet has doubled to almost 2 billion. The figures around number of actual sites on the internet are a little hazy, but according to Google the number of sites they had explored was 26 million pages in 1998, one billion pages in 2000 and ONE TRILLION PAGES in 2008, and the number of individual web pages out there is growing by several billion pages per day.
Now I know Google has some of the best infrastructure, technology and talent at their disposal but if we take the number of pages as products that Google has to provide quality control for, then that's alot of control - even for the world's biggest brand. Not only is there so much more information for Google to sift through, but the information is becoming more complex to evaluate. SEO is becoming more prominent, with alot of people who own web properties having been exposed to some kind of search engine optimisation education and using that knowledge to make their sites more 'appealing' to search engines. Popular blogging platforms like Wordpress also have SEO functionality now built in, straight out of the box, so again, more SEO-friendly pages are added to the web index. Aggregators, with their boilerplate templates are also becoming more popular in the SERPS again, further adding to the congestion.
Social media also comes into this. With Google and Bing starting to index Facebook and Twitter status updates, the amount of information is expected to increase exponentially. However, I think the search engines have the capacity to do this or they wouldn't try.
The Internet is expanding exponentially and pointless pages are taking advantage of Google's ranking system, rendering it unwieldy, Google and the other search engines need to counter this and find a way to once again become dynamic and streamlined and eliminate all those guff and filler sites that take up a large portion of their index (and my time incidentally).
Diminishing core focus on search
When Google first started, all they cared about was search. They put all their efforts and resources into making search better. Now it seems that have moved into fields pretty far flung from search, like renewable energy, cloud computing and mobile phones. Now I realise a simple answer is that they have invested in massive infrastructure and staff numbers to compensate for their diversification but somewhere along the way, with so many different things going on, their focus on search has diluted. Whether it's right at the top with Eric, Larry and Sergey or down the line through the many thousands of Google employees, it's happening.
I believe this dilution of absolute focus on search has resulted in some deterioration (or perhaps not as much improvement) of the search results. It's time to put the focus back on search, or the current batch of search engines may become the HotBot and AltaVista of the past.
Maybe search results have gotten better and we are just expecting too much, however I don't think this is the case; many people with in the search industry have noticed this and questioned the big wigs about it. I think that there are alot of outside factors that might be negatively affecting the SERPS, but I think it's time the search engines took cognisance of these factors and tried to counter them instead of putting feature upon feature (real-time search, more blended results, caffeine update etc) upon pretty average search results that are already crying out for some better quality control and more relevant sites within those SERPS.
Posted by Marc Elison on 14 January 2010 at 3:00 PM | Comments (0) | TrackBack (0)
See more entries in Industry News
7 January 2010
My 2009 online project
Photo by Titanas via Creative Commons |
Happy New Year to all. I hope your festive season and New Year parties were just grand. It looks like it's going to be an interesting year, what with the launch of the first Google (slightly HTCish phone), Facebook challenging Google for Internet dominance, and MicroYahoo challenging Google for search engine dominance. I'm not going to touch on any of these subjects, not now anyway, but they will be covered in good time.
Bruce Clay Australia has always fostered an entrepreneurial atmosphere within the office, offering ideas, support and encouragement for any semi-feasible web idea. Luckily enough towards the backend of last year, I stumbled upon such an idea. Now this isn't my first rodeo, I have built websites from scratch using good ol notepad, and (gasp) Microsoft FrontPage when I was just starting out. However this was my first experience using Wordpress as a CMS and not just a blogging platform. I have recently completed my project (no shameless link plugging here) and what follows are some interesting things I have learnt about the Internet, SEO and myself:
1. Expensive Wordpress themes are not always what they appear. I decided to use Wordpress for my project because it was going to be relatively static (aside from the blog section), and I am pretty familiar with Wordpress and if I wanted to do something, there would probably be a plugin for that. So I went about creating my Wordpress design. After about 3 days worth of work (with barely anything to show for it) I decided to leave it to the experts and went shopping for a theme. Awesome, after just a short while I found a theme that would suit my project perfectly, bought it immediately and set it up on my Wordpress installation. But wait, this doesn't seem right I thought as I stared at the screen that looked distinctly different to the demo of the theme I had just bought. It took many, many many, MANY hours of work for my theme to look anything close to the demo. In the end, it worked out as I got my theme looking just the way I wanted. But a word of warning to those of you about to purchase a Wordpress theme: most of the time the theme you buy isn't going to look like the demo you saw, so allow yourself a fair bit of time to get your theme right.
2. Double check your site is accessible. Now my site was finally finished, I had tweaked, proofread, optimised title tags etc, and I was finally ready to go live. I took off the disallow command in my robots.txt file and thought I was ready to roll. A day later, I feverishly checked my Webmaster tools and some accessibility tools and, to my surprise, my site was still being blocked to search engine spiders. So I started delving through Wordpress, looking for something I have missed. Sure enough, within the privacy section of Wordpress I found this little button ticked
I changed the setting and thought my job was done but apparently not because a few days later my site was still being blocked. More investigations for Marc. After going through every little link in my whole Wordpress dashboard, I finally found that my theme was blocking robots by default, so I changed that too and all was well with the world (well my website accessibility anyway).So in the end I had to check 3 places to make sure the site was accessible; the Wordpress option, the theme option and the robots.txt .This is a pretty vital step because you're in big trouble if search engines can't find your site. So always check your site is accessible and with Wordpress and themes it may entail you checking alot of areas around your Wordpress dashboard.
3. The All in One SEO Plugin rocks. I have always been sceptical of tools that claim to completely automate the search engine optimisation process as I have seen loads of these not come close to what a person can do. However, I had heard good things about this plugin so decided to give it a bash. Besides the fact that it also messed slightly with accessibility of my blog, it makes things alot easier from an SEO point of view. There are loads of features but the ones that made it useful for me were:
• Automatically optimises your titles for search engines and does a fairly good job of this
• Generates META tags automatically, especially great with the keyword tag as I know it's not relevant anymore but I have an almost sentimental need from the old days to always put these in
• Avoids the typical duplicate content found on Wordpress blogs, so you can very easily
i. Use noindex for Categories:
ii. Use noindex for Archives:
iii. Use noindex for Tag Archives
It doesn't automate everything but it does lend a helping hand.
4. The truth is out there (spoken to the background of the X-Files theme tune). So I ran into a problem that I just couldn't get right myself. You see, Wordpress has a nasty habit of placing the word 'category' or some category that you have chosen in the URL. Now I didn't want that, I really didn't want that. I wanted the name of my blog to be as close to my domain name in the URL as it can possibly be. So I typed "wordpress get rid of categories" into Google and voila, there must have been hundreds of articles and forum entries on the subject and plenty of walkthroughs to help me achieve this. Granted the first 5 little hints and hacks didn't really work but 6th time lucky I guess. The point I am trying to make is that Wordpress is so popular and has such a large community out there, that any problem or query you have, is likely to have been asked and answered before and you can use that solution to fix your problem.
5. Get another set of eyes. I spend most of my time on the content creation side of SEO. It seemed only natural that I should design a project with content as its focal point. So I wrote an ebook :) Now after months of creating, proofing and editing my ebook, I was quietly confident that there would be zero errors, typos and spelling mistakes. Just in case, I decided to get a few people I knew who had an eye for detail to give it the quick once over. My quiet confidence was a little unfounded: there were spelling errors, typos, etc. Now I am normally pretty good at that side of things but I think with this project I was a little too close to the content (almost knowing it by heart) and that inhibited my ability to be objective about it. It's cases like these where you need another set of eyes, just to make sure everything is ok.
It really was a learning process for me and being my own client also gave me an insight into what it's like to be a client and work with an SEO company. It's not always easy! I think that I have learnt a tremendous amount and with my future content creation, link building and social media optimisation, I doubt that the learning will stop anytime soon.
I just wanted to wish the readers of the blog a happy, prosperous and healthy 2010. I think it's going to be a good one.
Posted by Marc Elison on 7 January 2010 at 10:15 AM | Comments (0) | TrackBack (0)
See more entries in Industry News
10 December 2009
Google Search Updates
Photo by dannysullivan via Creative Commons |
Over the past week or so Google has released a slew of updates to its search operation. Some of these are somewhat controversial (personalised search) and some are just aesthetic changes (minimalist homepage) but Google seems to have ramped up their refinement process for the coming festive season. Here are just some of the new updates.
Google Personalised Search
From around the beginning of December, Google will start tailoring your search results based on previous browsing history, whether you are logged into your Google account or not(This review assumes that you are logged out). In order to allay privacy concerns you can opt out if you wish but most people using Google will be using personalised search by default. Google accomplishes this finding your favoured sites (indicated by which ones you most often click through to) and ranking sites from that brand or theme higher in your particular search results. This means Dymocks might be top of the SER's for the keyword 'books' while for another person the same search will yield Angus & Robertson as the top positioned site. This doesn't mean the other ranking factors are simply ignored however; they are simply used in conjunction with the personalised result factors. With this version of personalised search ( signed out of Google account), Google can only guess at your preference from what you click on, clicks are what your personalised search is based on, not search history or surfing habits.
Now you may think this has doom and gloom implications for search engine optimisation, well it doesn't, not yet anyway. First, Google is going to have to collect huge amounts of click data in order to understand and use your preferences within the SERPS. Good structured pages with terrific content, correctly used HTML attributes and links will still play the majority role in the ranking of sites, whether personalised search takes effect or not. We're not sure how much weighting the personalised search factor will have within the entire Google algorithm, but I wouldn't be surprised if people get near identical SERPS, regardless of their click history and habits. I have a feeling they may use the old Google 80/20 rule when implementing personalised search, i.e. 8 'normal results' and 2 results impacted on by personalised search for each SERP.
Google's New Search Page
Now unlike some of the other tweaks and updates that Google has implemented over the past month or so this is pretty much a purely aesthetic change. When the search page is first loaded, you only see the logo, search box and the search buttons. When you move your mouse the rest of the search page 'fades in'. This is something that Google has been testing for a while, for example it only worked on Chrome for a while and then periodically for other browsers. The only reason I can think of for this change is to funnel more people to start searching right away, rather than exploring the other links on the page.
Google Real-Time results
Google also announced that Google Real Time search has been rolled out. This feature is going to be rolled out in an incremental process over the next few weeks. Essentially Google is going to start indexing tweets, Facebook status updates (where made publicly visible by the Facebook user and this feature will be rolled out at a later stage) and other forms of real-time information. According to Google, they are processing over 'a billion documents a day' from the real-time web. In order to activate this feature you need to click on the show options link just above the results. This opens a whole heap of options, in order to activate the Google real-time results click on latest link. Then just sit back and watch new results blend into the top of your SERPS. Below is an example of my search for Obama, note the time next to each result.
Now with the advent of Twitter, 'real-time' seems to be all the rage, with everybody wanting instant gratification. I think the problem Google and the other search engines may be at the juxtaposition of real-time and relevancy. How do you rank a tweet in terms of relevancy to the subject? Is something that is 'real-time' better than something that is more relevant? Well, I don't have answers to these questions but with a couple thousand PHD's just sitting around thinking up answers to these questions im sure the big G has it covered. The following sources of real-time information will be indexed:
- MySpace
- FriendFeed
- Jaiku
- Identi.ca
- Fresh press releases or blog posts
Google Mobile Updates
Google announced a whole raft of search capabilities for mobile phones (and in particular android enabled phones). In terms of Google Voice search, Japanese, Mandarin and English are all now supported as well as several more phone models, including; Blackberries, Android powered mobiles and Nokia's.
Search By location is a really interesting feature. According to the Google announcement, people with Google Maps on Android 1.6+ devices, can just press on the map for an extended period and ten of the closest points of interest (restaurants, shops etc) will be revealed to you. Nice feature, especially if you are bored and stuck somewhere for an extended period.
Now this is a feature that really interests me, If Google can pull it off correctly it is going to be massive! This feature is currently a labs product that only users of Android 1.6+ devices can use. Essentially, you take a picture of an object, send it to Google and if they are able to recognise the object it returns relevant search results regarding the object.
According to Google, Goggles will recognise the following objects:
• Books
• DVDs
• Landmarks
• Logos
• Artwork
• Businesses
• Products
• Barcodes
Google has also released the following tips for making Goggles work better:
• Take photos in areas with good lighting.
• Zoom in as much as possible.
• Use your phone in "left landscape" orientation (with your right thumb pressing the shutter).
• Hold your hands steady, and use the on-screen shutter button.
Imagine the capabilities, you are in a foreign country, you are doing the
touristy/sightseeing thing, you want more information on an odd relic you are perusing so you take a picture and voila, all the information right on your mobile, very cool.
There are some other features like Google living stories, news rendered in a new and dynamic way. Updates to the Google Web Kit, including ways to speed up web apps. Loading time/web site performance is slated to be a big indicator in terms of search results in 2010. Martin has written a great post on Google and speed. As a result of the speed factor, Google has added another tool in Google Webmaster Tools where you can check the performance of your site. Google has also announced that extensions for Chrome are also now available. Wow, Google has been busy and I for one am looking forward to playing around with these news features and figuring out how they benefit me and the SEO I perform.
Posted by Marc Elison on 10 December 2009 at 1:15 PM | Comments (2) | TrackBack (0)
See more entries in Industry News
26 November 2009
Google on Speed
More Caffeine and more Speed
Google and Speed are very much interlinked. Matt Cutts officially announced that speed would play a role in the Google ranking algorithm. This announcement makes sense when put in perspective with the imminent release of Caffeine, an improved version of Google search with a strong emphasis on speed. The SEO implications are considerable, maybe more on Caffeine that on adding speed in the ranking algorithm. Marc wrote two months ago about Google Caffeine if you need more information on the topic. During the interview for Web Pro News, he released two key points about Caffeine. The first is that one data centre will be delivering Caffeine search results, representing a very small portion of the total search results delivered, and this will happen before the end of the year. The second point is that Google will start rolling out Caffeine across more data centres in January 2010.
Speed is already a factor in the AdWords quality score. So what is the speed factor going to change for organic results? Well, probably not much really, not in terms of rankings at least. It is more likely to give a small ranking bonus to faster websites than penalising the slower ones.
Why is Speed important?
The speed of your web pages can have a profound impact on conversion rates, on user engagement, and on the overall experience of your web site. When you have only 3-4 seconds to convince a user that you have what it's looking for, you don't want to waste half of this time loading the page, and run the risk of a new visitor landing on a page already frustrated by your site's user experience.
What speed is definitely going to change, is the overall online experience for the users. The improvements in Internet connections' speed in the past 10 years have increased the users' level of expectations and it now makes sense to improve user experience by preventing frustration through improved page load time.
More speed also means less bandwidth and hosting costs, more traffic, better conversion rates and therefore more profit for websites. It also means reduced costs for hosting providers. But the real winner is obviously Google itself. By encouraging webmasters to optimise the load speed on their websites, the world's number one search engine will benefit from less bandwidth consumption, less server space, better user experience, more AdSense ads impressions, and therefore more profit. This is without taking into account the savings that Caffeine's increased speed will also benefit Google, by increasing the number of search results and ad delivery frequencies (less load time = more impressions/sec = more click opportunities = more $$$).
What does it means for rankings?
According to the SEOMoz 2009 Search Engine Ranking Factors, "Registration and Hosting Data" is believed to only account for 5% of the overall Google ranking algorithm. it would be logical that speed would enter in this category, so it would be very surprising to see the share of this type of factors increase dramatically in the overall ranking factors.
However, a slight increase in search engine visibility is not such a bad thing for an SEO, and if faster websites might get a slight boost in rankings, no specific penalties have been announced for slower sites. The big question is to understand how they are going to benchmark your site's speed.
Some are already arguing that websites with better hosting solutions would be favoured compared to the smaller sites that only have a standard FTP hosting solution.
As the following study shows, most of the benefit to be gained comes from the frontend, which any webmaster can control, so I disagree with the ones arguing that companies with more money would have a competitive advantage when it comes to page load time.
Which Speed are we talking about?
When it comes to serving a page faster, there are two dimensions to look into, backend
and frontend.
A recent study by O'Reilly on website optimisation shows that a 50% reduction in backend response times (server side response processing, database processing etc), decreases overall times by 5% - 10%. Backend performance improvements also often require complex code changes, expensive hardware or software upgrades, which makes it more difficult to manage.
On the other hand, the same study shows that a 50% reduction in frontend performance decreases overall response times by 40% - 45%. This is the part where webmaster can have a significant impact on page load time.
The Performance Golden Rule
According to the Yahoo Best Practices for Speeding Up Your Web Site the Performance Golden Rule is that "80% of the end-user response time is spent on the front-end. Most of this time is tied up in downloading all the components in the page: images, stylesheets, scripts, Flash, etc. [...] 40-60% of daily visitors to your site come in with an empty cache."
Some of the good areas to look into include:
• Rule 1: Make Fewer HTTP Requests
• Rule 2: Use a Content Delivery Network
• Rule 3: Add an Expires Header
• Rule 4: Gzip Components
• Rule 5: Put Stylesheets at the Top
• Rule 6: Put Scripts at the Bottom
• Rule 7: Avoid CSS Expressions
• Rule 8: Make JavaScript and CSS External
• Rule 9: Reduce DNS Lookups
• Rule 10: Minify JavaScript
• Rule 11: Avoid Redirects
• Rule 12: Remove Duplicate Scripts
• Rule 13: Configure Etags
• Rule 14: Make Ajax Cacheable
Some Useful Tools
Google has published a very comprehensive list of tools
Some additional tools are:
Posted by Martin Orliac on 26 November 2009 at 12:31 PM | Comments (0) | TrackBack (1)
See more entries in Industry News
19 November 2009
Are Google spreading themselves too thin?
I know that seems like an odd statement for the most valued brand on the planet. However, I was just perusing the latest news from TechCrunch this morning and read 2 articles one after the other, the first was The Google Phone Is Very Real. And It's Coming Soon and the second was Google Holding Chrome OS Event Thursday. Complete Overview And Launch Plans To Be Revealed. Now I realise that there are tons of money to be made from alternative revenue streams, but is Google deviating from their core business and spreading themselves too thin? Are they moving TOO far away from their core business and cash cow i.e. search?
Danny Sullivan did a series of posts fairly recently that criticise the relevancy of certain Google Results and another post with a similar theme, that is The Myth Of Great Search Engine Results. Now I have to agree with Danny on this one, either I'm expecting too much or the quality of the SERPS is decreasing somewhat. You would think that with the recent competition from Bing, Wolfram etc, that Google would be piling their resources into improving and advancing search. They may be doing that in terms of engineers and PHD's but what about a guiding force? Most interviews or news by the big 3 of Google ( Sergei, Larry and Eric) involve wide ranging topics, from the purchase of a mobile advertising company, Google Chrome for Mac and the Google Books, not much on search.
However, I must cut Google some slack, with every advance they make in improving search quality, there are another 1000 spammers trying various very smart methods to game the system. There is also the ever increasing amount of sites being added to the web and therefore the index. Google also has to sift through these sites as well as older sites that are being given an SEO varnish.
There are many thousands of people at Google that know exactly what they are doing and I am sure they have 101 good reasons for diversifying ( not least, trying not to repeat Microsoft's example of not identifying a potential new technology and getting involved right away i.e. the internet). But are they going a step to far?
What do you think? Do you think that by Google diversifying, they are harming search, their product that made them famous (and billions)? Or is it outside factors like social media, exponential amounts of sites being added to the index and new technologies like Ajax that is making the SERPS's less relevant?
Posted by Marc Elison on 19 November 2009 at 3:59 PM | Comments (2) | TrackBack (0)
See more entries in Industry News
20 October 2009
Search Engines: Do as we say not as we do
Photo by Eliel via Creative Commons |
Well, I had the unfortunate case of waking up on this particular Tuesday morning with an affliction that affects all of us bloggers once in a while, the dreaded 'writers block'. While perusing all my online marketing related news and posts in my inbox this morning looking for some inspiration, I found an oldish article that certainly piqued my interest. It was a Search Engine Land blog post where they stated that Yahoo is indexing the keyword's tag after announcing that they don't ( in a follow up post Yahoo re-articulated what they meant, essentially they sent SEL a message stating "What changed with Yahoo's ranking algorithms is that while we still index the meta keyword tag, the ranking importance given to meta keyword tags receives the lowest ranking signal in our system", so they have corrected the mistake). However, it got me thinking should we as people in the SEO industry (or any other realm of online business) be acting on what the search engines say, or what they ACTUALLY do?
There is a whole lot of examples like this one though. One example that I find particularly amusing is that Google states "You should never have to link to an SEO" and you should avoid SEO companies that "puts links to their other clients on doorway pages". However, I have seen many SEO agencies rank for the highly competitive term 'SEO' based on just links (from what I can tell, although I can't say this for sure) from unrelated client sites (pages about dog food for example using 'SEO' as the anchor text and linking back to their SEO agency) along with 'SEO' in their URL and a strong age of domain as they have very little content on their sites and are still ranking very highly. This occurs across many different county TLD's. You would think Google would address that, especially as they maintain that links should be from pages focused on the same theme or basic content to have high value.
Another example of Google saying one thing and doing another was this example that was spotted by TechCrunch. Essentially Google changed the appearance of their buttons, and they did this by using CSS that "is not currently part of any standard and are only supported by Webkit based browsers (i.e. Chrome and Safari"). In addition the Google.com throws up 39 errors and 2 warnings if you try to validate its homepage for W3C compliance, especially strange as Google is a W3C member. Google accusing Microsoft/Yahoo of becoming a web monopoly also smacks of hypocrisy.
This may be a little bit of hearsay but digging up dirt on search engines is a little more difficult than I previously thought. According to this post Yahoo is offering a substantial financial grant to the University of Stanford for a programme of support for the journalists "working in countries not respecting the freedom of expression" all while working with the Chinese government to purge its china web index of 'undesirable information'. According to Reporters Without Borders, Yahoo.cn blocked a higher percentage of politically sensitive results than Google.cn or msn.cn (this was an old test so things might have changed), talk about saying one thing and doing another.
Now this is not meant to be a rant (although it might sound like it, it's not meant to be an attack on any entity) but I needed some examples that support my main point. The axiom I am trying to get across here is that you shouldn't believe everything you read on the internet, even if its source is one of the big search engines (who for some reason the majority of us trust implicitly) or a slightly nerdish guy whose surname rhymes with mutts and loves cats.
You need to read between the lines, listen so carefully so what the Google, Yahoo and Bing's of the world DON'T say as sometimes you can glean more information out of this, than what they actually do say. Rely on successful tests that you (and contemporaries that you trust) have run to evolve methodologies instead of Sergey Brin intimating that links don't have value anymore. Run experiments, collect data and make intelligent decisions on that data as opposed to hearsay. The SEO industry is one of the most dynamic in the business world and will only continue to grow, evolve and improve if we persist with challenging the conventional wisdom out there, finding new and better ways to get our clients websites in front of the right target market, so get out there and challenge & question everything!!
Posted by Marc Elison on 20 October 2009 at 10:25 AM | Comments (0) | TrackBack (0)
See more entries in Industry News
16 October 2009
Rich Snippets, Semantic web, Linked Data and SEO
You probably all have noticed that Google is displaying more and more rich snippets in its SERPs. These snippets include various information such as number of reviews, rating, price range, author, format, etc.
These rich snippets are made possible thanks to two major open standards called Microformats and RDFa that are now supported by the major search engines. In that matter, Yahoo has been leading the way since 2006, and we had to wait until May 2009 for Google to start supporting them. As far as I know, Bing hasn't made any official annoucement about supporting microformats yet, but seems to be supporting them somehow (see below).
Bing result for my favourite cafe in NY:
Same result in Google:
The microformats and RDFa are two elements of a broader project of semantic web called Linked Data initiated by Tim Berners-Lee. The basic principles is to mark-up the html using specific tags to make the information richer on the page.
The rich snippets are a direct consequence of this initiative, making these SERPs results more interesting for the users as well as for the webmasters who are encouraged to use them to gain more real-estate on the SERPs.
Three major types of information can be marked-up with Microformats: Reviews (hReview), Product information (hProduct )or Contact details (hCard). The same can be done using RDFa.
If you go and have a look at the citysearch page above, or at any Linkedin profile page, you will notice that the microformats or RDFa are being used in different orders, often mixed together. This is just to say this standards are flexible, and Google is getting very good at picking them up.
So what are the uses for SEO? Well for instance, you might want to mark-up the "Contact-Us" page to help the search engines pick up the contact details and improve local business rankings. When optimising websites selling or promoting products and/or services, you can use microformats to markup the product title, description, brand, price, review etc, or use them to markup reviews like most shopping websites already do.
Besides, some other standard are emerging such as the video microformats, and its RDFa equivalent based on the Dublin Core Meta Data Initiative. You won't be surprised if such initiatives start spreading for images, and think about the potential of marking up social media profiles.
Anyways, before publishing your page, you can use the rich snippet testing tool from Google. For instance, the test of the result above gives a similar result below, but when I tested my Linkedin public profile, the test displayed my location and position, but these are not appearing in Google.
One thing to bear in mind when it comes to rich snippets, there are no clear guidelines from the search engines, and they always mention that they "may" use them for their SERPs.
This evolution of the web to the semantic web will enable users to search through more and richer data, potentially being marked up by dozens of different tags referencing each others, inter-linking the web and therefore providing a new dimension called Linked Data.
The search engines are well aware of this, and the more they use these new dimensions, the more relevant their results will be.
Posted by Martin Orliac on 16 October 2009 at 2:46 PM | Comments (0) | TrackBack (0)
See more entries in Industry News
Marc Elison
Kate Gamble
Martin Orliac
