Forum Moderators: Robert Charlton & goodroi

Message Too Old, No Replies

The very very long tail at wikitravel

Learning from logs made public

         

stuartmcdonald

12:28 am on Jul 5, 2006 (gmt 0)

10+ Year Member



Subsequent to the buyout/sellout/merger of Internet Brands, wikitravel and world66 in April this year, Wikitravel.org made their logfiles available for public study. And while they were just basic webalizer stats, the search strings and referrers were very interesting to me.

The search string top 20 rapidly dropped off to the low 200s for particular phrases, yet with over half a million referrals from Google (all domains) this means there must be a long, very very long tail. Also interestingly they were ranking best for pretty generic travel terms -- I was surprised by that.

Please no stickies asking for the link to the logs -- it can be found easily enough by taking a bit of a wander through their site

BillyS

1:54 am on Jul 5, 2006 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Yeah, that is pretty interesting. With nearly 300,000 search strings in June 2006, the top 3 are basically looking for the site itself. After that I'm surprised too, they don't really seem to dominate a popular term, just a lot of terms.

tedster

2:40 am on Jul 5, 2006 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



Thanks for bringing up this topic, Stuart. This is the kind of long tail pattern I've found for many sites that you would expect to see reaping the traffic benefits of their top spot "trophy keywords". Instead, their business is really driven by an amazing diversity of longer phrases.

I see sites that have two big #1 results and that still only drives 5% of their business -- and yet management obsesses about that 5% and casually website elements that are having a much bigger impact. They will let their marketing department throw away a couple thousand legacy pages, just so they can "update their brand".

The long tail is not some kind of booby prize for hose who can't really compete -- in some markets, it's the most important action to be had.

suggy

6:18 pm on Jul 5, 2006 (gmt 0)

10+ Year Member



Make that in most markets. I spent ages building top level rankings on two really (100m-plus results) competitive terms for my business. Then I added a way to target the 'tail'. Needless to say the tail delivers ten fold the visitors.

europeforvisitors

6:34 pm on Jul 5, 2006 (gmt 0)



I've seen the same pattern. Diversity within a larger topic does pay off.

BigDave

6:37 pm on Jul 5, 2006 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member



On my biggest website I rank 1-5 for several of what you would think were the big keywords. None of those big keywords is in the top ten, ever.

What usually happens is that something is mentioned in a magazine and people search on it. Those are the ones where having a very long tail is helpful, since you never know what equipment the various magazines are going to write about.

The rest of the top spots are usually seasonal keyphrases or some of the slightly more obscure searches where there just isn't a lot of good information. For the big keyphrases there are usually a lot of sites with good info, for the smaller ones people might dig a little deeper.

On that site, I have not had a keyphrase that brings in over 0.5% of my search traffic in over a year.

I like the long tail.

jimbeetle

6:49 pm on Jul 5, 2006 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



The long tail is not some kind of booby prize for hose who can't really compete -- in some markets, it's the most important action to be had.

Yeah, if you have a halfway strong page in the first place and have taken some care in the writing to weave in solid related terms and all the usual stuff, well, you have an even stronger page.

I was just looking through some stats for the past few days. For a page that covers an event that happened yesterday, people found the page using 3,786 search phrases. I love the long tail.

europeforvisitors

7:07 pm on Jul 5, 2006 (gmt 0)



On my biggest website I rank 1-5 for several of what you would think were the big keywords. None of those big keywords is in the top ten, ever.

My site ranks #1 for any number of valuable two- and three-word keyphrases, and it also ranks high for some competitive single keywords. Some of those keywords/keyphrases do make it into my top 10, but they still represent a tiny percentage of total referrals. I'd guess that, on a typical day, my top keyword represents about 1% of my traffic from Google.

stuartmcdonald

3:46 am on Jul 6, 2006 (gmt 0)

10+ Year Member



One of the other interesting points is when you look at the terms in their June top 20 that are not generic (ie., the placenames) three of the four places are not what I'd consider to be "tourist hot spots" -- though rather oddly all are either in Indonesia or Malaysia.

When you search for better known tourist destinations -- say in Europe -- wikitravel results tend to be buried behind the plethora of other sites, so perhaps one of the reasons wikitravel has such a long tail is because they've got very thin coverage on a huge range of destinations nobody else bothers to cover, but for which a few people search every month...

ScottD

5:14 pm on Jul 6, 2006 (gmt 0)

10+ Year Member



we get loads of traffic from a whole variety of key words, which is of course wonderful, but I do sometimes wonder whether that traffic is useful - we sell widgets in someplaceland, so ultimately I suspect that the golden phrases related to that are the ones that actually give sales. Agree? Disagree? Or just still wondering where someplaceland is?

ken_b

5:32 pm on Jul 6, 2006 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



someplaceland

Someplaceland most likely has a nice set of long tails of it's own. So those are the long tails to work on. There will always be strangers who could care less about someplaceland wandering through, that's probably unavoidable, but what draws the strangers is what draws the someplaceland seekers even more.

To my mind, the beauty of long tails is that they can give a site a traffic base that allows a comfort zone while working towatds climbing up the more competitive terms for your market.

ScottD

5:41 pm on Jul 6, 2006 (gmt 0)

10+ Year Member



we are all over someplaceland with our content, but what I wonder is whether this is particularly good for selling the someplaceland widgets. Not everyone who is interested in someplaceland is interested in someplaceland widgets. I know it's not bad - but it requires a certain effort which may not be justified.

My point is the long tail needs to be of some relevance, or it is likely to be wasted effort, and if done badly could even put users off your site

ScottD

5:44 pm on Jul 6, 2006 (gmt 0)

10+ Year Member



To my mind, the beauty of long tails is that they can give a site a traffic base that allows a comfort zone while working towatds climbing up the more competitive terms for your market.

I agree with that. Also - I think the "weight" of the content pages adds to the PR / value of the site for Google, which helps the main pages to rank well long term. These pages will no doubt be aimed particularly at the main sales key words if you have a particular product, enhancing your like position for those golden phrases

webpro00801

7:03 pm on Jul 6, 2006 (gmt 0)

10+ Year Member



efv - has the latest Google shake up had any negative effect for you?

europeforvisitors

7:11 pm on Jul 6, 2006 (gmt 0)



To my mind, the beauty of long tails is that they can give a site a traffic base that allows a comfort zone while working towatds climbing up the more competitive terms for your market.

Just as important, those long-tail search terms are points of entry for visitors who will explore the rest of your site (assuming that your site has enough related content to interest those visitors). I have an editorial content site with a small number of affiliate-link pages, and most of the traffic on those very profitable affiliate pages comes from internal referrals--even in cases where the pages rank high in Google for fairly competitive terms.

ken_b

8:31 pm on Jul 6, 2006 (gmt 0)

WebmasterWorld Senior Member 10+ Year Member Top Contributors Of The Month



My point is the long tail needs to be of some relevance, or it is likely to be wasted effort, and if done badly could even put users off your site

I think this is where you get into thinking about negative keywords when constructing a page or editing an article.