Sunday, 23 November 2008

The End of SEO as we know it?

Google have recently graduated their social search product from Experimental to live for all Google account holders. Entitled SearchWiki, this tool allows the Google account holders to make changes to their search engine results pages (SERP's) that can be stored for access at a later date.

Here are my immediate thoughts:

How many people will use the tool?

It is estimated that 1.3M people in the UK have an iGoogle account. This can be used as an approximation for the volume of people that can use SearchWiki. However it is really only a small subset of these suers that are likely to use SearchWiki.

I use a lot of Google tools throughout a typical working week. I conduct hundreds of searches, am constantly using Gmail for email and file transfer, use Google maps to find directions to for meetings and football matches, analytics to monitor my web traffic, adwords to drive traffic to sites and webmaster tools to analyse technical details about my site. However, I can't realistically see myself using this tool.

The reason I use Google as my main search engine is because it provides me the best results in the right order. I have become accustomed with the how to search in order to give me the results I seek.

I also use Favourites fairly heavily and take advantage of the intelligent search bar within Firefox 3.0. Therefore, I cannot see why I would want to reorganise or annotate my search results as I can access the sites I need to in a simple way already. Is SearchWiki not just en elaborate favourites menu?

Will Google use any data from the tool?

At present, Google state that the information gathered from SearchWiki is not used to calculate organic rankings. My view however is that this will not be the case for very long.

Google are aware that organisations are currently able to reverse-engineer the algorithm and influence the organic search engine results. The principles of pagerank rely on users being able to hyperlink to and from websites. However, this means that search engine results can only be influenced by a small sample of Internet users that are capable of doing this. As the number of non-technical web users has grown, this percentage becomes even smaller. Therefore, it is important that Google take this volume into consideration in order to maintain the integrity of their search results. SearchWiki aims to achieve this by using Google users to shape it's search results.

Of course, it will not be long before SEO organisations adapt to start taking advantage of SearchWiki by using multiple Google accounts and increasing the rankings of their client sites.

I would love to know the level of adoption SearchWiki has seen since it's launch. Having been a fan of a lot of things Google have done, I struggle to see the value that making such a visible change to the SERP's would have on a Google user. I can only suggest that Google's intentions are not immediately transparent.

Sunday, 9 November 2008


Any web analyst knows that spending time evaluating trends in paid and natural search can add value by informing marketing and business strategy. Marketing directors are constantly asking what the most searched for keywords are and how this can inform their

Search statistics that are considered far less frequently but are potentially far more valuable is the results from on-site search.This is a valuable tool for many reasons.

1. Product Development Insight
If customers are searching for a particular product that is not sold on the website then this information is vital insight into consumer demand patterns. This can be used to inform product teams into researching the viability of selling that particular product. For example, a search for Flights to New York on the Easyjet website may give Easyjet an indication of the volume of customers that want to fly to New York with Easyjet. This is vital information that should be used to inform the product development process.

2. Customer Service
If it is clear that customers are asking particular questions via the search box, it may well be that there is insufficient information on the site. The help or FAQ section could be unclear or the consumer may not be given adequate information throughout their purchase. This information can be extrapolated to understand the volume of potentiual customers that do not complete a transaction due to insufficient information but do not use the search box. For example, if a large percentage of customers are asking about product delivery dates, it is worth making this more explicit throughout the transaction.

3. Merchandising Opportunities
The on-site search results pages provide a great opportunity to merchandise to engaged users. Ideally, an organisation would build an algorithm that responds to the users search terms with promotional offers and reasons to purchase. Failing this, advertising can be used to communicate key messages and promotions to users that are conducting searches on site.

4. Information Architecture
If customers are over-relying on the search box in order to find certain products then it may be that they are too difficult to find on your site. Thanks to search engines, web users are comfortable with using a text search box. But if the volume of customers conducting a search on a site is too high a percentage of total visitors, it may be necessary to review the navigation and hierarchy of information to ensure that the most popular products can be found as easily as possible.

The majority of web analytics tools are beginning to offer the capability of tracking on-site search. The Google solution even offers the ability to tie up on site search with Google Analytics in order to gain an understanding of the complete customer search journey.

Curry's on site search has not worked for me for a number of weeks now. Any search that is conducted on their site results in a frustrating session time out page. I believe that there may be a cookie issue causing session timeouts on their site but this fault skews their web analytics and upsets their potential customers. If there are people making decisions on this data, they could gain a very distorted view of how customers are interacting with their site.

For any decisions to be made on the back of on site search information, the first thing is it has to work.

Monday, 3 November 2008

Left vs. Right

Today sees the climax of potentially the most exciting and definitely the most expensive American election in history. It is widely expected that despite claims of a late comeback from Republican John McCain, Democrat Barack Obama will become the first black president in America's history.

The main reason for the huge cost behind this race, bar Sarah Palins wardrobe, is the unprecedented investment in media made by each candidate. Obama purchased 30 minute infomercials across several of the major TV networks last week

But how has this media battle played out online? Both of the candidates have their own websites which both feature number one in Google and Yahoo! for their respective names.
It is worth noting however, that both websites require Javascript in order to function correctly. This has accessibility implications, as visually impaired web users will not be able to access this content with their screen readers. I wonder how big a demographic, this segment represents?

John McCain's website has amassed an amazing 730,243 inbound links. In e-commerce terms, these are the sort of volumes that SEO's dream of acquiring naturally. This particular battle has a clear winner. Over the same period, Obama's site has gained 1,604,953 links. This is over twice the amount of his rival! Both sites have a pagerank of 8, a clear demonstration of the exponential nature of the pagerank algorithm.

But the online pain does not stop there for McCain. Running mate Sarah Palin has become a major character in the election race but in doing so has left herself open to attacks from viral networks, belittling her serious policies by concentrating on her image and appearance. In fact the Pew Research Center's
project for excellence in journalism found that 39% of news stories surrounding Palin are negative. This compares to 14% about Obama.

McCain has also had coverage of his political rallies removed from YouTube as he became another victim of the increasingly stringent copyright laws. As a result he only has 206K associated videos compared to Obama's 368K

So if the election was to be played out purely online then Obama is a clear winner. Although both websites have greater potential to further engage with their users and capitalise on modern online marketing principles.

Microsoft have developed an interesting visual search engine that allows voters to ask questions about subjects surrounding the election. It's not associated with any candidate or incredibly accurate but worth noting that this is Microsoft's third foray into the visual search arena. It's clearly work in progress.