Hitwise Intelligence - Robin Goad - UK
« Apple iPhone UK price drop | Main | Google delivers over a third of all UK traffic – trademark changes will have a significant impact »
April 22, 2008
London Mayoral elections: what does the Internet data say?
A lot has been made about how the candidates and voters for the London Mayoral elections on May 1st have embraced the Internet like never before - at least in British political terms, as this is one area where the US remains well ahead of us Brits. So, how are the candidates doing online? The first place to look is search data; specifically, how does each candidate’s name rank as a search term.

As you can see from the chart above, the Conservative candidate Boris Johnson was ahead last week, and has been for most of the campaign. Labour’s Ken Livingstone has been in second place, but last week was overtaken the Liberal Democrat candidate, Brian Paddick. Of course, it is important to state that searches do not equal votes – voters will be searching for their least favourite as well as their favourite candidate – but looking where people end up when searching for a candidate can be informative.

As you would expect, the top destination for each of the candidates is their main homepage, particularly in the case of Lib Dem Brian Paddick. Wikipedia is the second most popular destination, and is particularly important for Tory Boris Johnson. News and Media sites are most important for Ken Livingstone searches. As the incumbent candidate, a number of searches for the mayor also lead to the official London Government website. Charting traffic to each of the candidates’ homepages, you can see that the ranking for last week is the same as the search terms, although the trend in each case is more clearly upward, and the gaps between candidates are smaller.

Of course, not only do searches not equal votes, but website visits don’t either! However, looking at the story behind these visits provides some more insight. For example, 49% of visitors to Boris’s website live outside of London, while Ken’s site sends the most visitors to election / voter registration websites. This week the Labour party homepage was the fifth most popular downstream site from Brian Paddick’s homepage, while last week Ken’s homepage was the second – data that may have an impact considering the potential importance of second preference votes.
The most popular site visited before Brian Paddick’s site last week was Facebook, which brings us back to the original claims about this being an Internet election. As you can see from the chart below, Brian Paddick is currently receiving the most traffic from social networks, and this has played a part in the growth of his homepage over the last couple of weeks. Boris has gained traction on the social networks as well: his Facebook group is the biggest, while YouTube was the most popular site visited after his homepage last week.

However, when you look at the number of interactions across the social networks, they are actually quite small for an election involving millions of voters. The widely reported ‘I Think I Fancy Boris’ YouTube video has, at time of writing, received less than 22,000 viewings, while the collective membership of the three candidates’ fanclubs / friendship lists on Facebook is just over 11,000. Although the Internet has clearly played a role in this election, it looks like its still early days for online political campaigning in the UK.
Update (April 30th): With one day of campaigning to go, we have provided the latest data.
Posted by Robin Goad at 09:00 AM
Posted to Demographics | Government | News and Media | Politics | Search | Social networks | Video
Trackback Pings
TrackBack URL for this entry:
http://weblogs.hitwise.com/movabletype/mt-tb.cgi/786.
Comments
You say in your analysis of on-line trends that a higher proportion of hits are being made on Brian Paddick's website. I, too have visited this candidate's site but only by mistake as the feed did not mention that it was his address. I clicked on it purely because I hoped it might be an independant news story that could give me current opinion poll trends. Is it not likely that his lead on your graph albeit only of web 'hits', could be down to people like myself being led to a political site that is not titled as such?.
Posted by: Tim Hurford at April 22, 2008 10:03 PM
Thanks for the comment, Tim.
One way of checking this is to look at average session time, which for March was 4:04 for Boris, 2:47 for Ken, and 2:01 for Paddick. So, Paddick's session time is lower, but certainly not so low that those visitors are all arriving by accident. It does imply, however, that Boris's visitors may be more 'engaged' than the others.
Thanks, Robin
Posted by: Robin Goad at April 23, 2008 08:10 AM
Robin
I've just picked up that the Economist has dropped in Page Rank from 8 to 5 - looks like a text link ad penalty.
Could you possibly take a look at any traffic change there later in the week.
(Sorry for the off topic comment - I don't have an email for you).
I have commented in detail here:
http://tinyurl.com/6krlu8
Cheers
Matt Wardman
Posted by: Matt Wardman at April 27, 2008 02:22 PM
Thanks Matt - I'll have a look into that this week. One issue to conider is that the PageRank scores dispalyed in the Google Toolbar and other places for public view are not 'real time'. There is always quite a lot of debate ad to how up to date they are, and how often they are updated.
Regards, Robin
Posted by: Robin Goad at April 28, 2008 09:09 AM
I take the point about the occasion updates to the public databases.
I couldn't find anything in any published data - but it will interesting to see the results.
Matt
Posted by: Matt Wardman at April 28, 2008 11:23 AM
Hi Matt - I had a look through our data relating to the Economist, and I must admit that I didn't see any significant impact, either in term of the site's market share or the amount fo traffic that it receives from search engines.
Thanks, Robin
Posted by: Robin Goad at April 30, 2008 10:36 AM
