An SEO Experiment
First of all, If you’ve been here before you might notice that I’ve changed WP themes. The reason is that my previous one quit working for no particular reason (hacked?) and because I’ve been too busy (or lazy) I just turned on the default WordPress theme for a couple of weeks until I got around to dealing with it. Sorry about that.
The SEO Experiment
I have a couple of SEO questions that so far I haven’t been able to get a satisfying answer to.
- For organic search rank (not type in traffic, or navigational searches that include the TLD) does a domain name which is an exact or near match for the query phrase rank higher – all other things being equal.
- If the answer to that is yes, and the exact match domain is 301ed to another domain does it pass that higher ranking?
Because of some work that I’ve been doing I think that the answer to both of these questions is yes. But, my evidence is anecdotal, and I know that I could be wrong. However if I’m right this could be a useful tool.
So, I’m going to do an experiment to attempt to answer these two questions. Here’s what I have in mind:
- Obtain several domain names (at least 3 but more would be better) that consist of more or less random combinations of letters. Ideally they would be of equal length, bought at the same time and perhaps alphabetically sequential. I have to control the cost of the experiment so ideally these would be nice cheap .info domain names.
- Also obtain two or more .com domains that are an exact match for a bogus key phrase(s). The main point of this is to target valuable long tail searches so these domains would need to reflect this goal, thus – abc-defg-hijk.com or abcdefghijk.com although the distribution of vowels and consonants should be at least word like. There should be 0 results for a Google search of the bogus key phrase “abc defg hijk”.
- Set up bare bones web sites for each domain on the same shared host. While not ideal for real SEO this should help to prevent hosting from being a factor in relative rank.
- The web pages should be very simple plain vanilla html with no style formatting at all. Content should be extremely similar but not at all duplicate. I’m thinking something along the line of recipes with the bogus phrase being a key ingredient – not lorum ipsum gibberish. It needs to look like real content, and I’m thinking not optimized for anything on page.
- Submit site maps, and wait to be indexed. In the mean time don’t give any link love at all, and hopefully no one else would either.
Now my guess is that at this point the exact match domain would rank #1 and the others would fall out randomly behind it.
- Now 301 the exact match to one of the others and see if it now consistently ranks at the top.
Another experiment that could be done simultaneously with the same domains plus another (different) exact match domain would be to redirect it without it ever containing any content at all – nothing but an .htaccess file. Just to see what happens. I’ve actually done this, and it seems to work, but there are too many other factors to tell for sure.
I realize that a larger number of domains and repetitions of the test would be required for defensibly valid results, but this is what I have resources for, and I think that if it’s properly executed this could yield worthwhile results.
Questions? Comments? Predictions? Anyone see any problems with the method? Anyone know if this has already been tried?
Regarding SEO questions:
1. In my opinion – yes.
2. I have never tried, but I don’t believe 301’s pass on all the strength that a link has to offer.
Regarding Experiment:
1. A .info domain is inherently weaker in my experience a .com would offer the best test.
2. I would invent a word and then see how it runs – try “metatoastoptimisations” which can be broken down into “metatoast optimisations” and has only 3 other results to compete with. Plus you can then keep the domain redirected to you after the experiment so you benefit.
3. Set up bare bones web sites for each domain on the same shared host. While not ideal for real SEO this should help to prevent hosting from being a factor in relative rank. This may skewer the results – as they are all on the same IP – but maybe not. I have 8 sites on the same host all targeting the same keyword and they all appear in the first 20 results of Google.
4. This seems sound.
5. yep.
Finally I hope you will redirect the exact match to any domain other than the one that will be directly below the exact match result?
@rishil - thanks for the input.
2. I have never tried, but I don’t believe 301’s pass on all the strength that a link has to offer.
That’s part of what I would like to find out.
1. A .info domain is inherently weaker in my experience a .com would offer the best test.
I plan on using .coms for the exact match domains, but for the targets do you think it would make any difference as long as they are all the same?
Not exactly what you’re doing but here is what I have found:
I own the domain bowietownhomes.com. It was part of a grand plan I had a long time ago to subsegment my marketing efforts by town and housetype.
I then threw up a dozen or so links on some very small splogs (blogger blogs with only a dozen pages or so) and other places to bowietownhomes.com with the anchor text containing the words “bowie townhomes”.
Then I bought a very simple movabletype install for the domain. I never did anything with it. No content, no title, no optimization, nothing.
Check it out for yourself at archive.org. It was just something I let fall to the wayside.
It sat for several months and then I ended up 301′ing it to another domain - a basically empty Wordpress install with a theme copied from one of my other sites.
To this day that other domain “marylandwidehomes.com” ranks (#4 as I write this) for the search term of “bowie townhomes” (no quotes).
But neither the word “bowie” nor the word “townhomes” appears anywhere on the page or in the source code.
This has been the case now for months. I actually first discovered it through G’s Webmaster Central where they show you what terms you rank for.
Interestingly enough, if you search for the exact phrase of “bowie townhomes” (with quotes) - I’m not anywhere in the search results. But other much more powerful sites are.
For instance Trulia has a page with the exact title “Bowie Townhomes”. With everything I understand about large sites - this page should outrank my 301′d domain - but it doesn’t unless you search for the exact phrase.
If anything this is clearly a workaround to the google bombing issue where the phrase supposedly needs to appear somewhere on the page - you just need to set-up a keyword appropriate domain and 301 it.
I guess all of this is the flip-side of what you’re going for. I’m not sure how this would all work with an indexed site but no links.
@Vinny - Thanks very much. That’s some valuable input.
That strongly reinforces my suspicions, and backs up the anecdotal evidence I mentioned.
I had been trying to recapture a keyphrase that I had lost - valuable because of it’s high conversion rate. I registered valuable-key-phrase.com put it on an add on domain (on top of a different hosting package) with nothing but an .htaccess file pointing it at my landing page. Within 48 hours I’m back at #1 in the serps.
I’m still going to set up the experiment though because I’ve thought of too many cool things I can find out using a testbed of several identical (and yet not duplicate) websites.
I would be shocked at this point if there isn’t a domain-name effect, but it’s still an interesting experiment. I’ll second Rishil that I wouldn’t have .info for one set and .com for the other. I’d stick to .com for both to make this as applicable as possible to commercial sites (i.e. clients).
I would also plan on tracking the post-301 data at preset intervals (30 days, 60 days, 6 months, etc.).
Of course, the key phrase in all of this is “all else being equal”. Without links and other factors, I strongly suspect a matching domain provides an edge. What’s nearly impossible to tell, though, is how strong that advantage is. What if you tweak some keywords on the other page? What if one of the pages pick up a handful of links (even just 2 or 3)? It would be interesting, once you’re done, to see what it takes to offset the advantage. There’s a lot of SEO where we know (or very strongly suspect) that a certain factor matters, but we don’t really know how much it matters relative to other, similar factors.
@ Dr. Pete
I’ll second Rishil that I wouldn’t have .info for one set and .com for the other. I’d stick to .com for both to make this as applicable as possible to commercial sites (i.e. clients).
Agreed, although the cheapskate in me doesn’t like it.
I would also plan on tracking the post-301 data at preset intervals (30 days, 60 days, 6 months, etc.)
Absolutely, the longer it goes, the more accurate and useful the information would be.
What’s nearly impossible to tell, though, is how strong that advantage is. What if you tweak some keywords on the other page? What if one of the pages pick up a handful of links (even just 2 or 3)?
Because of all the variables involved it would probably hard to really quantify all of that, but we should be able to get a better idea. Once the experiment is going there are lots of tests that could be done using the same test bed of separate but equal domains.
Thanks for your input.
Something you haven’t mentioned above is what if you purchase a domain for the sole reason of using it to link to and 301 through? For example, purchase blue-widgets.com to link from your site CoolBlueThings.com through blue-widgets.com/competitive-term.html via a 301 redirect which lands them back on CoolBlueThings.com/content.html
I think that would be interesting to try and test.
The other SEOs have weighed in plenty on your original experiment.
Good stuff!
Thanks for the comment Brent. That is exactly what I’m talking about, but I’m wondering if it makes any difference if you redirect blue-widgets.com/competitive-term.html or just redirect the the entire domain?
Either way the original content on the original domain ceases to exist doesn’t it?
In the case immediately above - unless you have built huge numbers of anchor rich text links, the site will quickly drop off once the content goes.
BTW - if you are in a niche, then the actual page content is less important than your title, desc and internal anchors - vary those and you will see a diff, regardless of the content. Proven for my extreme niche sites. 7-9 results in page 1 and 2 for keyword for same client.
Rishil,
In that case do you think that the benefit of the the keywords in the redirected domain name is completely lost? That’s what I’m trying to find out (one thing anyway).
If you read Vingolds comment it would seem to indicate anecdotally that some benefit is retained even if there isn’t any real content on either site - not what I would have in mind to do by the way.
I dont think you loose all benefit - but a redirect benefit is lower - and I think its a reactionary tactic from Google when Bait and switch was a very popular tactic for link building.