Expedia's got no reservations on hotel analytics
By Jo Best
Published: 31 October 2008 13:04 GMT
Show relatedWhen it comes to online hotel shopping, there can be such a thing as too much choice, web travel company Expedia has found.
Given a choice between slicing and dicing search results to find the perfect hotel or just trusting the judgement of an online travel agent, most consumers will pick the latter, according to Dan Lynn, Expedia's director of strategy and customer insights.
Despite offering consumers the choice to narrow down potential hotel picks by price, star rating, distance and a number of other factors, most consumers are likely to pick a hotel from the first five results Expedia returns for their stay - a trend Lynn attributes to consumers reliance on search engines.
"One of our key challenges is inventory selection optimisation… The vast majority of consumers don't use functions [like sorting hotels by price or star rating] - they use the search results. It's critically important we put the best hotels for any query in the first few places," Lynn said, speaking at the SAS Premier Business Leadership Series event in Las Vegas.
The company has been using analytics to work out what attributes can turn a search into a sale.
"By looking at the data, running it through SAS and giving it to some great statisticians… we can figure out what the most important characteristics of a hotel or piece of inventory can increase the likelihood of that piece of inventory converting," he said.
"We can't get everything - we can't work out how the human brain works but…we can predict with 70 per cent accuracy what a consumer is going to select from a choice of hotels based on certain attributes.
"By taking each of these attributes and using SAS to build a choice model, we can score every hotel room we have and show which is most likely to sell to consumers," Lynn added.
But it's not just a question of tailoring those crucial first few search results to suit the customers needs, Expedia is also looking at optimising its site for a better effect on its bottom line - potentially by promoting hotels offering the best margin or simply better quality establishments to avoid a bad holiday for the consumer and encourage them to make a return visit to the site.
Another potential addition to the Expedia analytics armoury is behavioural analysis based on a customer's life stage - but the company says for travel, past purchases aren't always a good guide to future shopping.
Instead, Expedia is planning to work around clustering.
"What we want to do next is cluster modelling - cluster consumers into certain segments and then weight the importance of different attributes differently," Lynn said.
No comments:
Post a Comment