Thursday, August 07, 2008

What happened to the Long Tail of search?

Many of us will remember how, just a few years ago, many search engines and digital marketing agencies espoused the benefits of bidding for long tail search queries.

It was widely believed in some circles – indeed, some evangelised continuously – that users would gradually use a selection of keywords when searching, especially for travel products.

The theory went that search phrases such as “holiday rome”, for example, would be replaced by something along the lines of “city break rome pantheon april gatwick four star”.

There were some who even suggested long tail search queries would transform keyword bidding and SEO practices within two years.

So did it happen?

The MD of a London-based search agency, whom we met with yesterday, suggests it has not at all.

Consumers, our man with the data tells us, are refining their search phrase each time, rather than using long tail search queries to find specific travel product if they do not find appropriate results the first time.

For example: “holiday rome” becomes “city break rome” or “weekend away vatican” or “hotel rome weekend”. The use of lengthy and detailed phrases to hopefully narrow the number of results just hasn’t proven popular as many first believed.

What are the reasons for this?

Consumers may be sceptical that long, meandering search phrases do not return the required and, most importantly, quality information.

Secondly, users have perhaps found that refining the initial search with more detail is a far more successful method.

Or it could be something else entirely.

Indeed, these could be the wildly incorrect thoughts of a crazy digital marketer, so any additional comments readers have would be much appreciated.

Kevin May, editor, Travolution

12 comments:

Marco van de Kamp said...

I think most people like me think that when they use a search engine just use the precise words that describe what you are looking for. A sentence has a lot disturbance, words that not really matter.

If I like to stay a weekend in Paris with my girlfriend in a nice Hostel. I use google and search for "Hostels Paris" I use the websites that I get from google to search more specific.

Zorba said...

As is always the case with data it can give a point of view which best fits the deliverer, it would be interesting to see what sector of travel the data has come from i.e . bucket and spade package operator or a specialist operator?

From our own research the longer tail does seem to deliver a better conversion ratio than general keyword terms. This coupled with what I believe to be the growth of a more “general searcher” – which is derived from the increase in broadband uptake and less technically minded users who seem to treat search engines as a person; they tend to type more general terms like a “I would like to spend a week in Lake Garda in August” instead of “Lake Garda + August”.

A little more detail would probably shed some light on the findings…

Tao Tao said...

I think when you look at the successful or often-cited examples of the long tail such as Amazon, people already have a pretty clear picture of where they want to go, or as in the case of Amazon, the user is cleverly guide towards the things which he probably didn't even know that he wanted, but which are similar and cater to the individual's niche preferences (recommended products, similar products, etc.). I doubt that a completely open approach using just tags and search engines would really work, just yet. There's too much information out there, and often not relevant. Personally, I always stick to the websites that do what they do best, so if I needed travel information, I’d go to one of the platforms that is meant for it.

At GetYourGuide for example, we try to leverage the long tail by lettings users create tours and activities using google maps on a platform. Other users can then find, book and eventually review these tours. I think that one aspect of the long tail is that as a user you need a designated platform to access the long tail. In the case of books and CDs, this is Amazon as I would never start looking to buy a CD directly on Google. We try to be that kind of Amazon for tours and activities (without having to actually buy a whole package with flight and hotels).

Ultimately, I still think that users need guidance, or suggestions to put it more nicely. Maybe because they haven’t yet realised the potential that a more open approach could deliver or because most people simply don’t know yet where they want to travel before they start looking for it (hey, that’s where travel agents make their money…). Most people may think “beach” or culture”, but only a minority would really type in “tibet eco-tour lhasa monastery visit hiking” into google. Also, internet-savvy travel gurus may sometimes tend to overestimate the internet-savvyness on the supplier side of travel products.

Anonymous said...

I think of the long tail as stating that you get as much traffic from unique phrases as you do from your main keyword. For instance if I had a hotel with a blog in Pismo Beach California, I might get as many people finding me who are looking for night tours of Hearst Castle, ocean kayaking san luis obispo and california mission san luis as I do from hotel pismo beach.

On my site I am currently getting alot of visitors looking for info on the Dancing With the Stars Tour, Ryder Cup Golf and the Hike for Discovery. None of these are new posts.

Anonymous said...

One long tail keyword I noticed today getting a good number of visitors is, ‘cheap flights from Doncaster in august 08’ which sits in #4 along with Cheapflights.co.uk. Whilst this isn’t a very good example because I don’t offer, cheap flights, but, I get A LOT of traffic from long tail keywords that matches my niche.

So from my experience long tail keywords rock and drive great amounts of traffic but I think it depends on your targeted audience.

Anonymous said...

One point I've heard of recently to bear in mind, is that if you have a large number of obscure long-tail search terms in your Google Adwords account, and they are getting hardly any traffic, then this is going to impact negatively on the positioning Google gives you for your main search terms in the same account.
Can anyone confirm this ?

Travolution Blogger said...

All: thanks to all for your comments. i suspect this will run and run.

Yodey said...

Interesting views here. I became a bit obsessed with the long tail about 3 years when I set up Adprecision - with the sole aim of trying to get better quality PPC ads and clicks for the advertiser. Google is the de-facto top level generic search engine for the initial stages of research, once people have then narrowed down their search they will then visit a market vertical site like a shopping comparison engine, a brand owner (e.g. Hilton) or an OTA . I feel there is an opportunity for a very efficient travel vertical search engine that allows users to search in natural language with advanced filtering and navigation (because as we've learnt from Amazon / Ebay it's mostly about usability (hats off to Alex and Jonny at Want2bthere.com if they can get it off the ground).
We have also taken this challenge on board and created the AdTarget platform allow users to search by free-text for their ideal holiday and then filter using intuitive navigation links.

For example, go here and try searching on "4 star barbados holiday in october"
http://traveldirectory.timesonline.co.uk
Disclaimer - it won't give you a completely relevant result every time on every search you make, as we still have some way to go in convincing the travel industry that it's a good idea to make their product inventory keyword searchable. But we're trying!

Travolution Blogger said...

Webmaster: thanks for your comments.

further disclaimer: adprecision is running ad technology for the timesonline.

Londonsearch said...

What the Travolution Blogger says is true. As a large search marketing agency we generally run campaigns for corporate travel-related clients that utilise 120,000 to 1.2 million keywwords. What we've seen repeatedly is that for popular destinations, such as Paris, Rome, New York, etc., the long tail does exist -- though this is less the case when it comes to more obscure destinations.

Because overall volume for popular places is very high and general costs for converting popular terms are high, the savvy Internet Marketer will spread a wide net to lower costs.

It sounds to us like the MD Travolution blogger spoke to is trying to justify a shortened keyword list.

Tamara said...

HI Kevin
Interesting discussion. We find long tail keywords convert at a higher rate but there's just not enough volume for us compared with the standard 'hotels in Rome' type queries.
Personally, I do exactly as your MD said - I go wide first, see what I get then narrow it down each time until I get what I want. That's what I do naturally - next time I might try a long tail search just to see what happens!

Travolution Blogger said...

Interesting analysis here:

http://www.yucca.co.uk/blog_detail.asp?item=471