• All Analyst Weblogs
• North America
  • Bill Tancer
  • Heather Hopkins
  • Heather Dougherty
• Europe
  • Robin Goad
  • James Murray
• Asia Pacific
  • Tim Lovitt
  • Alan Long
  • Sandra Hanchard
• Insider
  • Client
  • Company
  • Industry
  • Product

Hitwise Intelligence - Robin Goad - UK

Analyst Weblog

« Vouchers vs. sales: battle of the online discounts | eMusic and the Long Tail »

Flight search and travel data update

January 15, 2009

After reporting a decline in flight searches last week in the FT and on our blog, we received a number of comments: on the blog post, via Twitter and elsewhere. Many of the questions have been addressed in the comments section of our and other blogs, but I also promised to provide updated numbers this week. So, here they are.

Firstly, data on the annual change in searches for our portfolio of ‘flights’ search terms:

• Week ending 03/01/09: -42.4% (this is the figure we published last Thursday)
• Week ending 10/01/09: -40.3% (new data)

The main feedback we received last week was that it wasn’t quite right to compare w/e 05/01/08 with w/e 03/01/09 because of the ways the days fell over the holidays – i.e. last year there were 2 more day post holiday days in the week. Now that we have a couple of weeks of New Year data, it does look as if the decline is due to something more than just a quirk of how the holidays fell.

Another way of looking at the data is to take the weekly changes in flight searches, this year and last, in order to see if the pattern has changed. As you can see from the table below, using the comparable weeks data from this year and last the pattern has been very similar. Traffic started to increase on a weekly basis starting just before Christmas and this trend has continued until now, although last week the rate of increase slowed (as it did during the equivalent week last year).

Of course, one theory is that people are booking flights later this year because they are still spending money in the sales or waiting to pay of their post-Christmas credit card bills. This may well be the case, and the test for that will come next week. Last year, the equivalent week (w/e 19/01/2008) saw the first weekly decrease in flight searches since Christmas. Again, I’ll be updating the data next week, as well as providing an overview of the January data in order to get a broader picture.

One of the other questions people asked was: given the increase in branded search (a trend that we have identified in the past), how much of this decrease in ‘flights searches’ is down to that trend? In other words, the correlation between changes in flight searches and traffic to travel / airline sites is not as strong as might be expected.

In order to test this theory we went through the top 500 search terms sending traffic to websites in our Commercial Airlines category for the 4 weeks ending 12/01/08 & 10/01/09 and categorized all that were either branded or contained an explicit reference to a destination (e.g. ‘cheap flights to spain’, ‘new york flights’, etc.). In order to understand the importance of these top 500 terms to the industry, they accounted for around 70% of search volume both this year and last. However, this year there were around 10,000 terms sending traffic to industry over the last 4 weeks, so we are certainly dealing with ‘head’ rather than the ‘tail’ here.

The first thing we noticed was that branded terms dominate the top 100, but as you get further down the 500 they become less significant. Still, there is no doubt the airline industry is extremely reliant on branded search: 95.7% of the volume of searches for the top 500 terms were branded this year, up from 94.4% last year. On the other hand, just 3.9% were destination searches, down from 5.7% last year (methodology note: we allowed double counting, e.g. ‘easy jet flight to spain’ would be tagged both ‘branded’ and ‘destination’).

Now, although this data doesn’t quite paint the full picture (the remaining 30% of search volume contains a much higher proportion of destination searches), it does look as if there is a natural decrease in the importance of ‘flights’ searches. So, does this mean there has not been an equivalent decrease in traffic to travel / airline web sites?

Yes, it does. For the week ending 10/01/09 UK Internet traffic websites in our Travel category was down 14.9% on the equivalent week in 2008; while for week before (w/e 03/01/09) traffic was down 16.3%. The equivalent figures for our Commercial Airlines category, which is a subset of Travel, were: -18.1% (w/e 10/01/09) and -25.4% (w/e 03/01/09).

So, in conclusion:

• Flight searches have continued to decline, and it looks as if the Christmas / New Year holidays fell did not have a big impact on this data.
• Traffic to travel and airline sites has also fallen, but not at the same rate as the decline in flight searches.
• The rate of annual decline in both flight searches and traffic to travel/airline fell last week. Based on this trend, it may be the case that the online travel season will extend further into January than usual because of the economic environment.

Stay tuned for a further update next week, and thanks again to everyone that has provided feedback,

Follow Hitwise UK on Twitter.

While we're on the subject of travel, here's some analysis of the Queensland Tourism 'best job in the world' PR campaign from Travolution, using Hitise data.

Tweet Posted by Robin Goad at 11:30 AM | (1) | (0)
In Categories Economy | Search | Travel

Comments