Why may much of what we ‘know’ about how guests select hotels be plain wrong?

I’ve been a reading over the last few months various studies that have researched how customers go about selecting hotels.  Much of this has been prompted by the significant increase in online reviews and the potential impact they may be having on hotel performance.  This blog is prompted by my fear that much of this research has lead to conclusions that are quite simply misleading – because of the way in which the research has been designed and the assumptions on which it is based.  So this blog is slightly unusual in that it is aimed at a very specific audience – hotel executives and professionals;  and because of this I am going to write it in a non academic style (so no sources, few stats, and bluntly).

An early example of how research in this field can get it wrong was in the 1990s.  Then there was a great deal of research into hotel ‘attributes’.  Studies often ranked these in order to show what influence these had on hotel choice.  The problem with this research design is that many of these attributes could only be known by the customer after they had checked into the hotel (e.g. size of the bath), so they could not influence choice.  Today this situation has been changed by online reviews, because they may well refer to these post experience attributes, albeit randomly.  Even so ranking attributes has a number of other problems as we shall see.

Partly because of this track record, in the early 2000s I supervised a PhD student, Maggie Chen*, who researched how customers selected a hotel using a travel search engine.  Much of my critique of other research below is based on what we learned from this study.  Unfortunately the results of her study were never published in an academic journal (otherwise the course of this field of research may have taken a very different turn).  In simple terms, 54 American students (not studying hospitality) were asked to use a search engine in order to select a hotel for a weekend break in Las Vegas six months in the future.  The process they went through was observed and recorded by Maggie.  Las Vegas was selected because it had a very large number of hotels (nearly 150) with vacancies on the specified weekend, so it was likely that subjects would use the search tools on the website in order to narrow down their choice.  Hence we would be able to see which criteria they used during this process.  In fact, one subject did not do this, she actually looked at all hotels – exactly how she then chose one, even she was not sure when asked afterwards.

One of the things we were most interested in is whether or not making the choice was a one or more stage process.  In other words did subjects use criteria to create a short list from which they chose, or did they go through this process a second (or even third) time.  For a few it was a one stage process.   But for most it was a two stage process – they used some criteria to short list (creating what is known as a ‘consideration set’), and then other criteria to reduce this list even further (creating a ‘choice set’), from which they then selected.  Now this is important, because some criteria were used to create the consideration set, and different criteria to form the choice set.  To have a chance of being selected, a hotel has to be in both sets.  But many other studies assume that choice is a one stage process, and they directly compare criteria that might influence consideration with those that influence choice.

I can explain the significance of this even further, by briefly discussing the actual nature of the criteria Maggie observed being used.  In simple terms (very simple – as we shall see), location and price were used to create the consideration set, whilst online reviews and ratings influenced the choice set.  So any study that concludes that price is more important than online reviews, or vice versa, is plain wrong.  Both are important in different ways – price gets you in the game, online reviews influence the final selection.

But all of the above needs to be placed in the context of what choice the subjects actually made.  Remember the sample was very homogenous – same age, occupation, nationality.  Therefore you might expect that their hotel choice would be homogenous.  In fact, 27 different hotels were chosen.  OK two were more popular than others, but the range of properties was surprising (to us at least).  Moreover, my simple summary of what criteria they used to make this choice is misleading.  The range of criteria used to form both the consideration and choice set was much wider than location/price and reviews/ratings.  In other words, a very similar group of people had very different criteria for selecting a hotel, which is why so many different properties were selected.  But many other studies are designed so that (a) they limit the criteria subjects can use, and (b) average the results in order to make statistical tests.  The trouble with this is that limiting criteria is obviously misleading; and averaging results only makes sense if subjects naturally cluster around a norm, which we suspect from our study they do not.

So what advice can I give to hotel professional on this issue?  First, be aware of and understand the limitations of any research study that investigates hotel choice.  Second, there is no one way that customers go about the selection process.  It is likely that even the same customer goes through different processes – depending on factors such as how close they are to a deadline, which search engine they use, how small or large your hotel market is, and their previous experience of that location.  Third, it is likely that a very high proportion of customers who use a website to select hotels will go through a two stage process.  Be clear about what criteria gets your hotel in the consideration set, and which get you in the choice set – for each ‘segment’ you serve (even if these segments may be difficult to identify in view of what I’ve just explained).  Finally, some of the criteria used in selection are ‘fixed’ (location for instance), so you can only manage those criteria over which you have control – price, prominence on search engines (maybe), online reviews (carefully), amenities.   And don’t even get me started on loyalty card membership!

* Maggie’s thesis is in the University of Surrey library.

 

Advertisements
This entry was posted in Chap 07 Capacity and demand, Sector: Hospitality & Tourism and tagged , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s