1. A data refresh – If you’ve managed to decrease the amount of spammy backlinks to your site, then great, that would be reflected in your rankings but if you’ve acquired more bad links or did not manage to get rid of those links which are hurting your website, then you will either be more affected or won’t see any change. External links are the hardest to control but remember over optimisation of On-Page elements will trigger a penalty as well.
2. Algorithm change – It seems the Penguin algorithm got better at detecting spam, well that’s what they claim. So new signals are being incorporated (not sure what they are) and their spam detection techniques have been improved. Hence they call it a major update.

Have I been affected?

Hell yes! A couple of sites I own have plummeted in rankings. I’d say the traffic have decreased by nearly 50%. Obviously I’m not happy about that but what is more frustrating is the fact I hardly build links to these sites anyway. Notice the word “hardly” which I’ve used and that’s because sometimes when I’m on a blog (with great content) which is relevant to the niche of my website(s), I tend to leave a comment on the page. More ever than not, I will enter my company/brand name rather than  my personal name but my brand name does not have any keywords in it. For example, the name could be something like “AirBubble” for a site which deals with web design, so there’s no way I’m trying to leverage the power of anchor text to increase rankings. My line of thought is to create brand awareness out there and in a month, I’d be lucky if I have been able to comment on 10 different blogs.

So what do I think about it all?

To me, it either seems Penguin is too strict or the people behind the algo have got something really wrong this time. Assuming there’s been no other algorithm change, it means all lost rankings can be attributed to Penguin itself. And if my websites are not acquiring many links (let alone spammy ones), then the algo is flawed.

It might still be too early to jump to conclusions but things are not looking right at the moment. Some SEOs have the tendency to blame others about being involved in black/grey hat techniques and hence they have been penalised but that’s not always true and if you’re one of them to think Google Updates are 100% accurate, you’re very ignorant. Many times, genuine websites get caught in filters just like when fishing for crabs, you might end up with unwanted fish and therefore it is foolish to assume a perfect win/win situation for everyone. The change might be correct for 95% of webmasters but that 5% innocent websmaters who got trapped unfairly can have their businesses ruined overnight.

My conclusion

Just like with any other Google updates, you will need to read and follow what other webmasters are reporting and try to locate the problem which is causing a drop in your rankings and find a solution for it. This is not as easy as it sounds unfortunately but the good thing about algorithm changes is that once you know what is triggering the penalty and can fix it, you’ll be released when the algo is renewed.

If your website has been affected negatively or you have benefited from this update, please let us know in the comments.

To explain the circumstances which led to this abnormality, let’s name the first webpage A (unoptimised) and the second webpage B (optimised). Page A has been around for over a year and was ranking for “aaa bbb ccc” and Page B was completely new but was meant to benefit from the domain authority as it was another URL on the same website (same sub domain). The keywords Page B was intended to rank for were “zzz bbb ccc” and as you can see it was only the first keyword which was different.

As both pages talked roughly about the same topic, Google thought Page A would be a better match for the query since it acquired more authority through aging and internal links. However Page A had only a mention of the keyword Page B was supposed to rank for and it was driving me nuts to see Page B not being shown in the SERPs but Page A instead.

This is when I revisited the term keyword cannibalisation and then it occurred to me the same thing was happening on my site. You see my title tag was so closely matched for both pages and Google was getting confused. The title of the webpage is very important and it was to be unique and you need to be targeting something specific with it. If you’re sending the wrong signal by having duplicate or near duplicate title tags, then obviously Google will have to make a decision about which pages it’s going to show in the search results and that sometimes may not be what you expected to see. This is how a non-optimised page was ranking instead of the most relevant page on my domain, the one which talked in depth about the subject in question.

To overcome the problem, I’ve had to find a variation of the title tag which was a bit longer but included the keywords I was aiming for and guess what, 3 days after making the change, I was happy to see the intended page was ranking instead of the other one. The reason why this is important is because if the visitor does not find what he’s looking for when he lands on your webpage, he will hit the previous button immediately and you will have a lot of bounces. He doesn’t want to see an article where what he’s looking for has been talked about briefly but wants something better than that. Also your conversion rates will go down if you have a non-optimised page ranking.

So it is in your best interest to have an seo audit and check your website properly for places where you might be cannibalising your keywords and apply the remedial solutions straight away. The good thing about this is that what once you identify these pages which are affecting your conversions, you can quickly re-rank them again. In my case, my rankings have been stable for more than 7 days now.

I have used the data from CleverDodo.mu from the last 6 months (October 2012 – April 2013) and found 10% of visitors to the website accessed it through a mobile device. Here are the top 25 phones/tablets:

1. Apple iPad
2. Apple iPhone
3. Samsung GT-I9100 Galaxy S II
4. Samsung GT-I9300 Galaxy S3
5. Apple iPod
6. Samsung GT-N7000 Galaxy Note
7. Samsung GT-S5360 Galaxy Y
8. Samsung GT-I9300 Galaxy SIII
9. Samsung GT-P7500 Galaxy Tab 10.1
10. Samsung GT-S5300 Galaxy Pocket
11. Google Nexus 7
12. Samsung GT-S5830 Galaxy Ace
13. Samsung GT-N7100 Galaxy Note II
14. Samsung GT-I9000 Galaxy S
15. Samsung GT-S5830i Galaxy Ace
16. Nokia C3-00
17. BlackBerry 9300 Curve 3G
18. BlackBerry 9800 Torch
19. BlackBerry 9900 Dakota
20. BlackBerry 8520 Curve
21. Nokia 5800 XpressMusic XpressMusic
22. SonyEricsson LT18i Xperia Arc
23. Asus TF300T Transformer Pad TF300T
24. Nokia C5-03 C5
25. Samsung GT-S5570 Galaxy Mini

It is amazing that many Mauritians own an iPad which accounted for 31% of mobile visits while the iPhone was 23%. People have started to move away from desktops and laptops and are adopting tablets and smartphones more nowadays. You can’t blame them though as mobile phones can do pretty much anything these days, from making a phone call to taking high resolution pictures and there are plenty of apps for just about anything else.

Looking at the data from a popular UK website I own, I saw the figure for mobile access to be much more (37%) but that’s no surprise at all because smartphones are much cheaper in England and internet data plans are more affordable. The top 5 mobile devices for the UK was:

1. Apple iPhone (45%)
2. Apple iPad (23%)
3. Samsung GT-I9300 Galaxy S3 (3%)
4. Samsung GT-I9100 Galaxy S II (2.8%)
5. Apple iPod (1.2%)

What these figures suggest is the iPhone is more popular in the UK while the iPad in Mauritius. The research also shows the need to have a responsive website design so that the layout is resized according to the screen resolution of the device accessing it to provide a better user experience for your visitors.

The main reason for sandboxing would be to fight spam. If you had a burst pipe in your house and you had to choose between 2 plumbers, one who is completely new to the business and the other who’s been in service for 5 years, you would without a doubt choose the one who’s been around longer. Similarly, a new website *would* not achieve its full ranking potential until it has been properly vetted by Google (or other search engines).

To further illustrate my point, let’s consider a website which offers medical advice. If someone was searching for some sort of rare disease and came across this newly registered domain which contained a lot of information about this, that person may believe what’s written is completely true and may try whatever solutions were listed on the webpage and cause harm to themselves. Unless someone in the field can vouch for the accuracy in the information provided, you shouldn’t trust the information.

Spammers (people who do not contribute to make the internet better but looking for ways to make easy money) don’t want to spend the time and effort to build a successful website which stands the test to time. They just want to make cash fast and they’ll create lots of websites to make them profits usually at the expense of quality. To combat this problem, the sandbox inhibits the ranking of new sites. An aged site is not one which has been registered a long time ago but one which has been first indexed ages ago. For example, you might have had a domain for 5 years but you only posted something 2 years ago and Google crawled it 2 years ago. So really, your website age would be 2 years only. Now an aged website does not mean better rankings, that comes with quality of content and trust your website has built over the years.

How much are the rankings held during the sandbox?

You shouldn’t really expect a lot for the first 6 months. During that time, it’s best to concentrate on adding useful content to your website, branding it and getting your site socially connected and engaged. You can refer to the following table for the sandbox effect with time:

Month    Holdback

1                100%

2               75%

6               50%

12             25%

24             10%

36             5%

48             0%

I’d also like to mention that creating a sub domain on an aged domain does not guarantee top rankings. The sub-domain will be treated as a new one and will enter sandbox and therefore it will have to create its own reputation as well. The only benefit it can have from being linked from a popular domain is that it will be indexed faster and can be transferred some link juice as well as getting a strong signal that’s the sub domain is important. This reinforces the rule of impeding a website from ranking well initially to prevent spammers from gaining first page results by registering a new domain and building spammy backlinks to it.

UPDATE : I created a new sub domain on 28 November 2012 and it would not rank even though it was getting editorial links from some authority websites. Surprisingly enough, today (24 May 2013), a little more than 6 months later, it’s ranking on Page 1, 5th position for its main keyword on Google.mu. This shows new websites (domains and sub domains included) go through that sandbox phase during which they need to build with Google before they start ranking.

• Open 24/7 – You’re no longer restricted to 9-5 working time. Your shop remains open to your customers every day of the week for 24 hours, including public holidays.
• International market reach – You are no longer restricted by geographic location. You can now target customers in other parts of the world with your website.
• Cut down on costs – You may not require as many staff or even a actual shop to run your business. You can trade only through the internet if you want.

Besides those listed above, with an internet business, you can offer multiple payment methods, easily sell to your existing customers (send product/service update through newsletters), get statistics on what people like most in your shop and easily add offers/incentives to attract new clients.

I hope by now you’re already convinced how having an online shop can drive more sales for your business. So let’s see the steps involved to get your e-commerce shop off the grounds:

Picking up the right domain name

Many people ask me which domain they should buy (.com, .net, .mu etc). I strongly advise a .mu extension if you’re going to be doing business primarily in Mauritius. This will not only geo target your website but will also have a high impact in the SERPs (search engine result pages which are the pages you see when you search something on say Google.mu). It is a known fact that a country level domain has a much higher CTR (Click Through Rate) but the major drawback with .mu domains is they are very expensive compared to .com – 4-5 times more expensive actually. Once you’ve settled on the domain extension, it is time to choose your domain name and if you don’t already have an idea what your business is going to be called, I suggest finding a name that you can use as a brand rather than something tacky.

Building up your website

Next you need to find a web design company to actually build your e-commerce platform. This includes creating a design which fits with your business image and doing any programming to accommodate the functionality you desire in your website. Note that you will need a web host where you’d be able to put the contents of the website so that people can browse to it through the Internet.

Choosing the payment methods

There are many payment options for ecommerce solutions but not all of them is applicable for Mauritius. For example, Paypal is not supported in our country, therefore you’d need to find another online payment gateway such as MoneyBookers (Skrill), 2CheckOut which are well popular on the international level or choose MCB (Mauritius Commercial Bank) or SBM (State Bank of Mauritius).

There is usually a setup fee, a monthly charge, a transaction fee as well as a fixed percentage on the transaction amount which will be charged for the Online Payment Solution. So make sure you do your research well before signing up for the payment gateway.

Adding your products to the database

When your ecommerce website has been setup, you would need to enter all the products you want to sell in the database. This tends to be a tedious job, so if there’s any way to automating the task, you should go for it especially if you have hundreds of products. Talk to your web design company to see how they can help. Remember to put as much information as you possibly can about the products you’re selling as well as clear pictures and videos if you have so that the customer can make an informed decision straight away when landing on your site.

Marketing

The last step is the actual marketing of your website. It will be useless to have the most attractive website if no one knows about it. So notify your customers and spread the word amongst your friends and family and watch the orders coming in. You may also be interested in PPC (Pay Per Click) or Social Marketing (Facebook, Twitter etc) to get more exposure for your business.

Other things to note

You might require a trade permit from your local council (municipality) to run a business either from home or somewhere else. So make sure you contact them beforehand. Some products require a special permit when importing; eg flowers, plants need a permit from the Ministry of Agriculture and other items may require approval from the Ministry of Health. Therefore do your homework well before indulging in any kind of business.

Something to be proud of…

We recently (January 2013) launched a brand new ecommerce website for the Clever Dodo Boutique which aims to make online shopping for Mauritians easier by providing a service where you can buy and pay for your products through the internet and have your order shipped over to you anywhere in Mauritius. The website might come in handy to you, so don’t hesitate to check them out at shop.cleverdodo.mu. You never know, you might get addicted to it!

Otherwise, if you want us to design your website, then get in touch with us now for a quote.

Anyway, some site owners have tried everything they could to get all their bad links removed but they as expected, not all of them were successful in having 100% of the low quality backlinks taken off. With the introduction of the Google Disavow Tool however, a webmaster is able to make a text file containing links they would not vouch for, that is, links they would not want their site associated with. You would need to have  a Google Webmaster Tool account where the site in question has been verified. You would then create a text file (name it whatever you want) and add the bad links in it. An example of such a file is as follows:

http://badwebsite.com/badpage.html

domain:spamwebsite.com

As you can see, the contents of the text file contains 2 entries. The first one is a specific URL and the second one tells Google to ignore all backlinks from spamwebsite.com. I would advise you to use specific URLs instead of banning whole domains because Google would need to recrawl the webpages where you have the backlinks before flagging them as being ignored for link calculation. If you were to use the domain operator, it would take much more time for Googlebot to find the page where you have the bad link.

Another point worth mentioning is that the disavow tools serves as a strong signal to Google just like the canonical tag which means, it is really up to Google to use this preference/indication you’ve given them. In most cases, they would use it though but if they believe you’re shooting yourself in the foot, then they might ignore some/all links in your disavow file. Please also note that you would need to wait first for Google to recrawl all those links you’ve specified as not wanting to be used for link analysis and then wait for another Penguin update to see whether it works for you or not.

I’ve myself tried this with an exact match domain (EMD) that contained 60 odd links from article directories out of a total of 70 backlinks. If that works, it’s going to be really awesome. I also think that Google will use the contents of the disavow tool to better understand what sort of sites are spammy in nature. For example, if a lot of webmasters are putting the same site (spammy.com) as an entry in their disavow file, then surely Google will treat spammy.com as spammy due to the sheer amount of votes. This can also mean competitors can list your white hat site in their disavow file and get you penalised. Might need to dig more into this but later…

Organic traffic is basically visitors that are sent to your website when they use search engines. Say you’ve written about “where to get the best deals for laptop” and someone is searching for that and Google thinks your webpage is a good match for the user’s query; it will therefore refer the user to your site and that’s a good thing. Search traffic is how many people get their sites known on the web and it is important that your website is properly optimised in that respect.

However misconfiguration or inadequate knowledge of SEO can often lead to loss of potential rankings. Say for example you have a page which should really be ranking #1 for a search term but you are currently on Page 5 (#48). You are missing out a lot of this search traffic. A deep analysis of your website reveals that you have a lot of duplicate content on your website which has caused a penalty that you were not aware of.

This happened to one of my clients. She was using Blogger for her site and had written 10 blog posts to help people understand the benefits of her product. When I investigated, Google had indexed 90 pages from her site that was showed there was a problem. What happened was that Blogger uses a feature called “Labels” which are tags to associate words with the blog post. For instance, after writing a post on “how to fix a leaking water tap”, you could tag it with “tap”, leak”, “fixing tap” etc. This resulted in several pages with the same content. In a Post-Panda world, this is an absolute NO-NO. Duplicate content should be removed.

Using the robots.txt file to tell search engines which part of your site they can index

Blogger already has a default robots.txt file for you but it was mis-configured. They know that search/label pages do not add any value to the blog and they have specified in in the robot exclusion file but the entry was wrong and therefore Google was interpreting it as an invitation to index. I created a custom one for the site with the exclusion as follows:

User-Agent : *

Disallow : /search

This basically means that I don’t want any search engines to index any part of the site which starts with the word “search”. So /search-me.html, /search/, /search/labels/, search/dummy/content.html would all be excluded.

If you choose /search/, then everything within that folder will be excluded but not /search-me.html.

You can test specific URLs of your website against your robots.txt with this tool.

We’ve also prevented archive pages from being indexed by adding a bit of code to the blogger template. This should help the site rank better for the keywords they are targeting.

If your robots.txt is wrongly configured, that could mean access to whole your website could be denied too and you wouldn’t appear on Google search results.

Although free CMS are great at first glance, it is important that you check or have someone who knows their stuff to check whether you are not inhibiting your own website’s success.

PS: If you’re using Blogger or WordPress, you should make sure that your post titles come before your blog name so that you rank better.

So the colleague came over and I asked what SEO services their company was offering. He told me that they only do search engine optimisation if they build the website themselves. He said that the structure of a website is really important and that’s what matters to search engines like Google. Granted a good structure is important so that navigations are user friendly and the content is organised semantically. However this is only part of SEO, referred to as On Page Optimisation and the other more challenging part is Off Page Optimisation and that’s where link building and social media promotion comes along.

Since I wanted to know about the techniques that the company uses for SEO, especially link building, I told the chap I wanted to optimise a website of mine and would not mind having them redo the whole site if that’s what it takes for them to do the work. Confident of himself, he told me we can know instantly if a website has been well SEOed and he took a title from a webpage on my site and google it. To much his disappointment, he found the webpage was indexed. He tried another 2 webpages and reluctantly said that my website is already well optimised. I told him the on page side of things is fine but I wanted to rank better for the keywords I’m targeting.

He told me everything revolves around good content and he’s not totally wrong. However if you and your competitor both have extremely good content, then how are you going to outrank him? I believe SEO is the answer for that but the chap was not convinced that 2 people can have equally outstanding content. His views on SEO is that you should write good content and that’s it; it will rank by itself. This is loads of crap. Honestly! I’ve seen website with really outstanding content that are either not ranking or are being outranked by some other witty publisher with a moderate piece of information on the same topic. This is all to do with quality link building and actively promoting your content on social media platforms.

Our friend from the well known company doesn’t have a clue about SEO and yet is offering such services to the public. Many people do not really understand the basic concept behind search engine optimisation and they may be tempted to go with such companies because they’ve already made a name. However, you will no improvement if anything at all regarding traffic to your website because to me the person is a conman.

A good search engine optimiser must be able to tell you how he’s going to go about doing link building for you, where he’s going to get the links from and roughly how long it’s going to take for you to achieve top rankings. If they say they can’t disclose their “secret sauce”, then they are obviously hiding something and most of the time it’s not a good sign. Although no one can really guarantee top positions in Google, an SEOer must be able to increase your website’s rankings gradually through safe practices. Note that in 2011, Google released a Panda Update and in 2012 the Penguin Update and if the person you’re hiring to bring in more traffic to your site does not have a clue what those 2 updates are, then I’d advise you to run a mile because not only will you be wasting your money but your current rankings might be affected as well. Many people have used SEO services and have suffered great loss because their rankings have either disappeared or they’ve lost significant amount of traffic due to grey/black hat techniques.

As you can see from the above picture, a wordpress administrator bar is displayed at the top of the homepage (and all other pages). This is common when you’re logged into your WordPress blog as admin because it makes administering the site easier since you can be browsing your site/blog and the admin options are just readily available to you. However for all other users (especially non-administrators), this should be hidden or made unavailable.

The “Clear Cache” button on the wordpress bar was very prominent and when you click on it, it was performing a cleanup on theindependent.mu. Only the administrator of the site should be allowed to clear the cache but here, anyone was allowed to click on the button and subsequently getting rid of the webpages which have been cached. The purpose of cache is to make webpage loading faster. So for a dynamic website where there is a round trip to the database to fetch records (data), it is more efficient to get the data once and store it a static page instead of doing the same routine of fetching each time the webpage is requested. In this instance however, this is not working as expected and the benefit offered by the cache plugin will be useless if someone were to hit that clear cache button ever so often.

From the wordpress bar, there’s a wealth of information made available to you. For example, it was showing that 62 comments were in the moderation queue and that there was a newer version for 9 plugins that the site is using. Using outdated plugins is a big NO-N) because it poses a security risk and that can get your website hacked easily. If people know you are using an old version of a software or plugin, they can find the weakness in those scripts and use that to gain illegal access to your site. That’s why it is really important to stay up to date with new versions of applications you are currently using.

Of course, all the admin functionality was inaccessible because it required you to log in but this does not excuse the ignorance of the webmaster. The problem is that the cache plugin that was used was either wrongly configured or poorly designed. A good cache plugin should not cache pages when you’re logged in because it will confuse users making them think they are already signed on when they are not. More dangerously, it could give access to sensitive parts of your website to a casual visitor who unintentionally can bring havoc to your whole site by clicking on things they don’t understand or just out of malice, try to do things to damage the site because the opportunity was presented to them.

The purpose of this post is not to embarrass anyone but to show that negligence can have really devastating effects. Security is of utmost importance and testing is a necessity when you change something in your website. This could have been prevented by configuring the cache plugin, logging out and testing whether the functionality is working as desired, logging back in and testing once more and then making the changes live. Hopefully they will correct the problem before something bad happens.

Therefore Facebook login is a really good way to register users on your site. You don’t have to send them email with a confirmation link to verify their accounts because Facebook has already done the hard work for you. With a simple click, users can be registered on your site and post comments or have access to any other functionality that would otherwise require a registration. This works well since people hate having to register on every single website they go to (fill out the register form, choose a password, check inbox to verify email address etc) and having to remember yet another password.

As a webmaster, you’re happy because your users are happy with a simplified process to interact with your site. You are able to get their email address and name as well (other details could be asked for t00). However the problem arises when you get an invalid email address with Facebook Connect. Let me explain that because it won’t make sense until I do. To register on Facebook, you need a valid email address and upon registering, you will have to confirm that your email address is genuine by clicking on a link that Facebook will send to you. Some people however, stop using that email address with which they registered on Facebook. Many free email service providers like Yahoo/Hotmail will delete your account if you do not check your email account in a period of 6 months. If that happens, then when someone tries to login with Facebook on your site, you will eventually get an invalid email address and that’s where the problem starts.

The whole point of using Facebook Connect is to make it easy for people to interact with your site. You therefore don’t have a way to verify that the email address you’re getting is valid or not. Well, you could actually send the normal confirmation email to the user but to me it’s an overkill. Fortunately most people who use Facebook have a valid & active email address associated with it. So the chances of that happening is really low but still it does happen and I found out about this because notification emails were being bounced back to me for some users who have used Facebook login (Remote host said: 550 Requested action not taken: mailbox unavailable).

On a side note, Facebook does its best to alert users when they are using an email account which is invalid. When the user logs onto Facebook, there’s a message telling them to update their email address as they’ve been having problems delivering emails for notifications (likes, comments etc). Should the user not update their primary email address, then the problem will continue.