Search marketing: April 2008 Archives

Currently in beta, Firefox 3 has introduced some fab new technology to the browser location bar.

Users can perform a keyword search through their entire web history from within the location bar.  This makes for a fast and convenient way to navigate back to pages that have previously been visited.

The technology goes way beyond auto-complete by offering a keyword search of URLs, page titles and tags from previously visited pages.  Results are ordered by 'frecency' - a combination of how recently and frequently you have visited a page (with an adaptive learning algorithm further tuning results for relevancy).

For example, let's say that you have been browsing through Tesco Direct and found a Whirlpool washing machine that you like.

You decide to check back with your partner before making the purchase. 

Autocomplete on the URL isn't too helpful... you'd have to know that the URL http://direct.tesco.com/q/R.200-9621.aspx related to your chosen washing machine.

But type 'Tesco Whirlpool' into the browser and Firefox locates the page you want...

Normally this is the kind of heavy-lifting job that people would have taken to a search engine.

Last year, Atlas published an analysis on 250,000 clicks on paid search ads on Google Adwords, Yahoo! and MSN.  The report, 'Paying for Navigational Search', showed that 36.9% of clicks were searches on brand terms by repeat visitors to a website.

The big brands moaning about Google's relaxation of brand protection rules in the UK should be praying that Firefox 3 really takes off - the potential is here for companies to get chunks of relevant traffic direct from the browser, just as nature intended!

Our friends at Insiders View have an interesting post about Google's 'previous query technique'.

This has been in varying degrees of beta testing for at least six months, but now seems to be happening a lot - particularly around high cost financial services terms.

What happens is that Google will take a user's previous query into account when serving results for a search.

So here's an example.  My first search was for 'car insurance' and brought up the usual suspects.  I then followed up with a search on 'Paris'.

Here's the resulting page:


To be honest, I think that this is a bit of a mess from a user's point of view.  The top three sponsored links relate to car insurance, but everything else on the page relates to Paris.

Are all these links relevant to the search query?  Not really. 

Of course, Google would argue that the links are relevant to the searcher, who has indicated an interest in car insurance and in Paris. 

But what's really going on here is that there is a limited amount of good financial services inventory, so Google is 'spreading' the search results into other pages.  It's exactly like the way that retargeting on some portals and networks operates, where someone who has checked on loans offers can be shown a loans ad whilst they check their horoscope.

And as a comment on the Insiders View blog points out, this is yet another attempt by Google to grub in a little more cash.  So if you reverse the searches, and look at 'Paris' and then 'Car Insurance', there are no listings at all relating to Paris, France or anything other than car insurance.  Google's results are not following relevance, they are following the money!

Google's first quarter results were spectacularly good and certainly blew away any thoughts of faltering performance from our favourite search monopoly.

Deep in the numbers is the performance of Google UK. 

"Revenues from the United Kingdom totaled $803 million, representing 15% of revenue in the first quarter of 2008, compared to 16% in the first quarter of 2007 and 14% in the fourth quarter of 2007."

So year on year share of revenues from the UK have dropped by 1%.  But that reflects greater earnings growth in other less mature markets rather than any inherent weakness in the UK.

A year ago, Google's Q1 2007 results showed UK revenues of $578 million.  My maths is a bit shaky, but that looks like a fairly healthy rise of around 40% year on year.  An ordinary company would be pretty happy with that, although it's a little less than the global increase in revenues of 42%.  Over the past 12 months sterling has appreciated a little (around 1%) against the dollar, which takes a tiny bit of shine off the numbers, but they still look very healthy.

All the same, Google seems determined to extract as much revenue as possible from paid search in the UK.  Hence the impending change to bidding on trademarks, which is likely to drive up bid prices on brand terms for most large companies.

Last August we also saw some changes to the top ad placement formula on Adwords.  Instead of looking at actual CPC when calculating the cost of a top ad placement, Google now looks at the maximum CPC that an advertiser is willing to pay.  This will add to the pain of brand owners who will naturally be keen to keep their own brands at the top of the paid search listings.

Finally there is a slow drip of changes to what Google seems to regard as legitimate SEO techniques - a post on seobook (Google Thinks YOU are a Black Hat SEO.  Should You Trust Them?) sums up the current state of play very nicely.  It might be a cynical way to look at the market, but it looks as though Google is saying that if you can afford to do SEO on any scale, you should be pushing your budget into Google Adwords.

Google enjoys an incredible market share in the UK and all the signs are that it plans to maintain strong revenue growth off its monopoly position.  And no one would bet against them achieving that!