What Would You Do If It Did?
Recently I found an exact match domain for a little project, with 4 domain extensions. I popped a little bit of text on each so as to be ‘unique’. A one page holding site for each. I left for Google to discover.
A few weeks later all 4 exact match domains are in Google(.co.uk) and ranking for their term in the same vertical (with another 168,000 results).
- .com 5
- .net 12
- .org 21
- .co.uk 23
Your rankings of course, will ultimately be determined by your content and incoming links, and the rankings will fluctuate, but it struck me as slightly interesting to see the difference in ranking between the sites, as I have often wondered where randomness factors into Google – if it does.
In the test sites, the titles are the same, the keyword is mentioned the same amount of times etc etc… theres only 50 words on each page max. There really is not much different between the pages – at all – apart from the domain extension.
If you have a .com in this case, you are laughing – immediately in a top 5 position. But if you choose a .co.uk, you start from the 3rd page? Dead in the water. At least, your starting from a different point.
Perhaps it’s to do with the domain extension, but perhpas it is an indication of how Google works at a granular level – the discovery phase – perhaps at this level, your positions are assigned randomly based on a particular set of principles (which we will never know).
Perhaps this randomness is prevalent in Google inner workings and is what protects it from us ever finding out exactly how any particular element works, and even employees knowing it all, or being able to ‘promote’ -
Matt Cutts did say on his blog:.
someone walked up to me and pretended like he wanted to bribe me: $500,000 for a 1st place ranking. I turned him down, because no one can guarantee a #1 ranking — not even me.
I’ve REALLY tried to isolate some fairly simple elements use in the past – some I thought MUST give me a definitive answer but alas they did not.
If this was the case it means trying to actually figure out how Google works is a non-starter – it would mean there was no sweet spots, anywhere. Perhaps it’s different for all sites. For all elements. Join that together with some ranking elements that are turned OFF, or tweaked, personalisation, geolocation etc etc and you have something that can’t be gamed. Well, too much.
Perhaps this randomness is more diluted for the top sites, than the churn they sit on (everything after page 2 or 3)?
What would you do?
It’s actually very easy to get good returns from Google unpaid listings if you give it what it wants.
You can stack the odds in your favour by adding lots of content and getting credible links to your site. That’s what seo is for me. Look at what the competition is winning with and try and figure out how to 1. compete and 2. beat them. Usually that means copying them to a point, and then trying to do something better at some point when inspiration hits.
Google has a lot of spammy verticals it seems to want to let you spam your way into them as long as it’s a good relevant site which is at least as good as the current competition that have already. Of course, some verticals seem more protected than others, but that could just be the level of competition, or an ‘age’ thing.
It seems as if Google purposely uses brands to clean up verticals with lower quality competition. If I see a vertical with a lot of big powerful brands as the top ten I think ‘hello’ – here’s a vertical Google needs some help with. Brands (well, specifically internal pages, like a bbc article for instance, are good, but they don’t beat focused anchor text linkbuilding on their own). Or even a great exact match domain that’s been in a low quality linkbuilding campaign.
To get the most from any element, you probably need to be a player – an online entity – a site with trust in any type of competitive vertical.
Which means getting BETTER, or more trusted, more credible links than the competition has, if your page is RELEVANT.
Google is clearly going to be using signals from brands for a long time to come. Links from online brands will make your website a brand until it finds another way of finding trusted sites.
I see a ‘brand’ as a real site, with some real links to it (or fake real links). This is probably why the seo companies who put links in their client websites rank at the top of the SERPS. I don’t ask any seo clients for links, but I ask folk we’ve made websites for, for the odd link.
As soon as Google can access your pages, with simple navigation, with original content, with a good title – it really is about getting links from real sites.
This is all pure theory – just a mind wander if you are into seo geekery. Don’t go changing your domain name or anything silly. maybe this is all just a cae of – it looks like that – sometimes.
The point of my article I think is to point out even though you don’t have all the answers, getting quality links is and will be the most important thing you can do to get better rankings from google. If you want more traffic from Google meantime – add content. Lots of it.
Shaun, Which version of Google did you search for the rankings US or UK? The low UK placement could be a country relevancy issue? This post exemplifies what I love about your blog… Good, clean original SEO geekery. Some variables I’ve thought about while reading this post: 1. Maybe you were ranked by different robots. I’ve often found I do well with new content that will quickly die out. – It will be interesting to see how these sites rank in say 3 months. 2. Maybe Google found “relevancy” in your 50 word gibberish. 3. Perhaps the rankings reflect when Google found the sites, or the source they first crawled each website from. Be interesting to try this experiment again. Since you posted about the science of SEO recently, how would you set up a control group for this experiment?
I am in the UK, accessing Google.co.uk (all the web) – so that would be counterintuitive looking at these results.
Never really thought about that.
I doubt it for these differences…..
Yes, I think about this aspect in previous ‘tests’ (used loosely). Still, the differences are a bit much based on crawl time difference of a few days/weeks at most.
Hmm – you say you looked in Google.co.uk…. so I really hope it is down to some of the issues suggested by Portland – like the order in which the sites were found etc. Because for Google.co.uk to rank a .co.uk so poorly is really bad/worth considering… Lx
Alright Shaun, I’m still in the baby stage when it comes to how long my website has been around and I don’t understand some of the SEO geekery as you called it. However I do understand a lot more about backlinking methods since coming aboard SEO Hobo. I’m happy to say it works for me. Mom and Pop always said,”If it ain’t broke, don’t try and fix it.” Thank you and God bless.
Thanks Randy – I should update those books they are a bit ‘unprofessional’ and perhaps (gasp) a bit out of date. Still, doing the basics right is where it is all at :)
Hi Shaun, Agree with where exact match domains automatically sit in the rankings. I find exact match .co.uk domains with 1 page of content always go straight in at page 3 or 4 – only needs a few links for a page 1 position. You do need to add further content though as they often drop off a bit.
Yes, Carl, and that’s what I do as soon as there is a site developed on there :)
Were all the sites hosted on the same IP address, this could have made a difference. I am really surprised to see the .co.uk ranking last on Google UK.
Yes – interesting note perhaps that could have impacted things. Next time I will use different IP :)
Do you know whether there was any previous use of the .com domain? Out of the four, probably the most likely to have previous owner. It’s definitely something that should be balanced out – it is very difficult to rank in the States for a .co.uk, but fairly easy to rank in the UK for a .com Perhaps they think that a brand is most likely to own the .com for it’s site and therefore this is weighted. Interesting post for discussion.
I can’t see any previous use of the domain at all.
Hi Shaun, Another great post that really makes you think. Do you have the luxury of other sites with a similar spread of TLDs to test with? That would have to be the way to test this and see if similar results come up. You would have to check the progress of each site over time too to determine any longer term effects of the TLD.
I will probably do the exact same thing again – I might try to control it a bit more next time.
I’ve been working as a consultant for an international hotels group and have noticed a distinct US bias towards domains. In particular, a Spanish hotel appears on both the US .com site and the UK .co.uk site, yet despite the UK site being closer to the hotel, better optimised, bigger, more content and more recently updated, the US site is completely dominant for almost all keywords. In contrast, Bing and Yahoo show the UK site as in the top 10 — and usually the top 2 — for almost all the same keywords, with the US site being nowhere. Both sites had similar backlink profiles (i.e. none) and link-building has since brought the UK site to the top half of page 2 on Google (and still rising) however, across the group’s websites this .com bias is still very clear with even UK hotels losing out to their US .com counterparts, so this cannot be a country-specific issue.
Great post Shaun, I always thought that the local TLD (.co.uk) will be promoted on the Google.co.uk search. If the sites are hosted on the same hosting server, it might be that the geographical location of the server may have something to do with it. Overall great observation.
Hi Frans thanks for the RT Indeed. But, the sites are hosted in the UK too ;p Of course, I say in the article, I would expect the rankings to fluctuate. I am more interested in where they started from. :)
Nice one Shaun. This just brought back me to a long dead topic in my head. For a while i thought if having different extensions on a domain name would help me get a better place on serp’s. For example just like co.uk in UK, we got com.tr in Turkey. For a while i was thinking of getting the com.tr domains so i can get better serp’s in local search, but as you said it took me a while to figure that some of those sites were being placed randomly before they received any links whatsoever.
Great post, I’ve been toying with the idea of forking out whatever the current owner of the respective .com is for my site. Interesting to know that there seems to be a difference. I doubt I’ll bother shifting from the .co.uk to the .com though.
“As soon as Google can access your pages, with simple navigation, with original content, with a good title – it really is about getting links from real sites.” This basic formula does seem to be enough for most verticals. We still find that implementing best practices gets many clients to the front page without any heavy lifting. One comment in regards to getting quality links. We have a client with what I would consider a powerhouse link profile. Only issue is none of the links are from relevant sites. (thx to the services of their prior SEO who promised them first page in 90 days) By not relevant I do not mean generic directory listings. More like a dentist with tons of links from great lawn equipment sites. So I’ll assume by “quality” links you include the relevancy factor.
Er, no. I would take a top quality link on a powerhouse site in an unrelated niche (newspaper, news site) before any relevant site. It works for me. I always try and get some relevant links over the piece, but I would rather have a link from the BBC. Where’s the relevance there? Yet a link from the bbc can put a page to the top of Google – I have a few examples :)
Shaun, I know of sites that haven’t been updated in years, that are not the best in their vertical but the shear number of backlinks of varying levels of “quality” allow them to rank higher than the better, fresher, content filled sites. While no one knows Google’s algo, there are times I have to wonder if lots of quality, fresh content is actually what, as Google states, they are looking for. I discovered your site about a month ago and now find I look forward to seeing “Hobo” in my inbox. It is sooooo refreshing to read your commonsense down-to-earth approach. Keep up the great work!
Hi Judith Adding content to your site won’t necessarily improve your rankings for a key term, but it STILL is the number 1 way for most to get MORE traffic to their site. For instance, my rankings for SEO has never been worse. But we probably get more seo related traffic than most other sites in this industry in the UK. :)
The weirdest thing about that to me is that the .org ranked lower than even .net.
Interesting experiment. I’m also curious to know if any of the domains had been used before (to see if age was a factor), and if so, are there any incoming links from the previous use.
[...] Does Google Play Loaded Dice With Your Rankings?, Hobo [...]
You mentioned that the sites were hosted on the same IP, but where was the server located? I’d expect a co.uk site hosted in the UK to perform slightly better on google.co.uk than a .com site hosted abroad, although obviously it does depend on a lot of other factors.