A bit of of seo testing :)
A lot of people use gibberish words in tests so I don’t. I use unique numbers which only appear in hidden elements – and I have a lot of pages all over the place. I am only interested in what I can see and what actually helps a page rank in SERPS -
- Q: Does Google, Yahoo or Bing use the meta description as an aid to rank a page?
Does Google Use The Meta Description When Ranking A Page?
My test page comes up No1 for the keyword search. THE QUERY search result is here. It’s no1 – note that.
I have a unique number in the meta description which is not on the page.
If I search for – “THE NUMBER” + “THE QUERY” I get no results – Google says there is no single page in it’s index with those 2 terms on the page, even though there is. Query here.
These tests have been live for months and at least in this qualitative test, it certainly appears as if:
- Google ignores the meta description when ranking the page for the query - if keywords are only in the meta description and not on the page, the page gets NO Benefit in a real life search.
Google knows it’s there – you can check that using the info: operator with the same “THE NUMBER” + “THE QUERY” search and it’s clearly visible in the SERPS when the page IS retured.
BUT THIS ISN”T HELPING A TARGET PAGE RANK – AND THIS IS ALL I AM INTERESTED IN.
Does Yahoo Use The Meta Description When Ranking A Page?
Yahoo DOES use the Meta Description when ranking a page it seems….
Does Bing utilise the meta description in this way? It does not return the page. Bing seems to work the same as Google – effectively ignoring the meta description tag according to this test.
Takeaways
- Google is ignoring the meta description when the keyword is not also present on the page – which kinds of suggests Google ignores the meta description WHEN RETURNING A BASIC SEARCH RESULT TO A USER.
- Google still reads the meta description and uses it in snippets if you are using certain operators – if you want the meta description to appear as the snippet, place your keyword in it (and that keyword needs to be on the page, too).
- Yahoo does seem to still use the meta description when returning a page in this manner
- Bing is similar to Google (there’s a surprise – sarcasm) – it does not apparently use keywords in the meta description that are not on the page
It also might mean Google wants us to think it is ignoring the meta description totally – certainly, this is a good way to nulify the unique word in the meta description test, and any seo tester testing on a granular level will tell you, Google likes to nulify or muddy these kind of tests.
Which kinds of makes it hard to test things on a granular level these days.
Don’t trust me though. Test yourself, on the site you want to rank. You just need to ensure the keyword is unique to the page, and not in links or mentioned anywhere else on the site. Which is kind of hard.
Questions This Test May Not Answer?
Does Google reward a VERY RELEVANT meta description? Again, very hard to tell for sure (I’m running tests on this too).
Run your own tests and see what you can cup with, but for me, I still just try and make good pages that are as relevant to the phrase I am targeting – and IMO, you can probably always make a page more relevant. Sometimes I include a meta description sometimes I don’t. Usually, I write it for humans first, then engines.
I just thought it was worth sharing for discussion as I seem to use a slightly different method of determining things to some of my seo buddies with tests here and here.
Do add any comments or let me know if you can think of testing – I use that word loosely - things a bit more accurately.
I’ll share some more test results for discussion later, too.
[...] This post was mentioned on Twitter by SEO in 1 Place, Shaun Anderson, Matt Rycroft, Matt Rycroft, SEOidiot and others. SEOidiot said: RT @Hobo_Web: Meta Description SEO Tests – Will Google, Yahoo or Bing Use It For Ranking Pages?
Thanks Shaun, I really appreciate you sharing your methods, results and conclusions. There’s value in this and I find it compulsively fascinating too! :-) Mark P.S. got your Algerian flag ready?
Try this query – http://bit.ly/cyCXYf and u find hobo in there
Alan – Yes I know, that’s proof that Google looks at the Alt text of an image. More on that later :)
Come on then Algerians lol ;)
This is great and reinforces another study I saw a few years ago. However, there is one tiny flaw in your logic. Just because Yahoo is returning a result based on the meta description doesn’t mean they are using it as a ranking factor. You’d need two pages with the unique item – one with it in the description tag and the other not and you’d have to completely control for all other factors to determine if the description tag had any ranking power. Nice work though Shaun. Thanks for sharing.
I’d like to see the test again, comparing one site with the number just in the body and one with the number in the body and description and see if the one with the description ranks more higher than the other. This would suggest that with competition, Google DOES use the description, even if it’s just a slight amount. What do you think?
This is a good example of how things are moving forward in SEO. I have heard this theory before and to be honest, it does make sense for Google not to use the descriptions as it is an easy target for spammers. However as you rightly say, the descriptions should definitely be written for humans in order to encourage them to click-through to your site. The more people clicking through to your site is bound to have positive impact on your rankings so in a roundabout way META descriptions are still important.
cool post, thanks for the suggestion got a few ideas I fancy trying out now. Thank you for sharing.
hi Shaun, I love your posts where you supply actual TEST RESULTS to back up your ideas. Keep it up! As to this one, I have thought for a long time that google ignores meta content except to use for snippets. Unfortunately, people sometimes see the meta tags used for the snippets and then think that they help for rankings too. As your test shows, this is (most likely) not the case. Great post! Steve
Testing search engines can be difficult if not impossible! If you use a real world example, Google, Yahoo or Bing (if they thought it was insightful) could just ammend the test results as to what they want you to see – or nuke the pages/sites in question. In this case, it’s hardly search engine terrorism. I think it’s more useful to see the actual test pages in these cases so you you can make your own mind up – so I decided to share a way of looking at trying to determine if Google counts the Meta Description when ranking a page as there is some recent discussion on it in a few places this week – and indeed better help formulate an even better test if there is one :)
Hello Shaun, First off let me give you a huge thanks for taking the time to provide this informative and educative resource. This website (blog) is top notch and it gets straight to the point, in addition to your clean layout and design. Second, It seems meta tags are very important but also not important. (oxymoron?) LOL. They only play a part when the on page SEO is correctly optimized. For example. a website, actually lets say webpage, that includes Keyword A in the title, in the content, AND in the meta description should definitely get more “juice” from G rather than a page that only places Keyword A in the Meta description and NOT in the content. I don’t mean to sound redundant but I just wanted to be sure what you said. Is this correct? Thanks again for your hard work. I’m now a frequent visitor of your site.
Hey Mike :) Generally speaking I think you are quite correct. Don’t expect to rank for keyterms in ‘invisible’ elements. AND – always try to make your page as relevant as possible for a query or range of queries. That’s my opinion at least.
Just had a look at the results in Google and by simply plugging in the page, i.e. 9583769348576348, a single result is returned in the SERPs. I’m sure you’re going to dig into this at some future point (if you haven’t already done so) but it does verify the importance of keywords in your domain/page title.
Yes James I aim to write that up too :)
Thanks for the post. Got some nice ideas to try out. Thanks for sharing your results.
Hi Shaun, Your experiment was so interested for me and I decided to test it by myself. As I’m not in UK but in Bulgaria I actually translated your content and upload it on my site. The only question I have is how to tell Google to index the page? The easiest way I know is to point that page but then I’ll give the page more ranking as it will have a link form a high rated site… Will this brake up the result? And what was your way to tell Google for your test page?
Lovely Test Shaun. I concur with Tradeshow Ninja; It’s lovely when you actually get results after an important test, and you share it with us :) We already know that Google already ignores the keywords tag. I wonder if they will start to ignore most of the meta tags from now on. But the test actually makes sense. I don’t think i would rank any page for a term either (emphatise search engines ftw!) if it’s included only in description and not in body text. This would mean that the word is a random one and not an important keyword :P
Cool – does that mean I beat you to the punch on that one? ;)
That would be a no lol Why do think that number is in there? It’s a example page – it actually tests quite a few things….. ;)
Ah well, no kudos for me today :(
Very cool and interesting tests, I have been wondering the same about Yahoo and Bing, always knew Google never gave an edge to it. Thanks for the testing results, shared via FB!
I have to say that I think that the test results make perfect sense. Why would a search engine add a page to a search query if the keywords in question never appear on the page. Good call on Google’s and Bing’s behalf on this one. Interestingly, this makes me question YAHOO if they list results based on the meta description. Keep the results coming, Shaun! You have gained one more loyal reader…:) I just wish I could subscribe to new posts via email…:)
Thanks for your testing efforts. It’s good to know that at least Yahoo uses the meta description. In a Google Webmaster Central Blog post from September 21, 2009, Matt Cutts admitted that Google doesn’t use meta description data in their rankings.
Hey Dave, Yes I was aware of that :)
[...] it’s probably worth paying attention to your websites meta description, although as this experiment suggests, you should ensure any keywords included in a pages’ meta description can also be [...]
ei Shaun, I always like your write-ups, particularly your sharing of info on every experiment’s results and feedback. And as a newbie on SEO, every article on this blog is priceless! Thanks! btw, I’m a subscriber of Hobo for more than 4 months now, however this is my first time that I actually commented here, if am not mistaken….. dannngg rss! :) Kudos to you! ariston
I have to say that I think that the test results make perfect sense. Why would a search engine add a page to a search query if the keywords in question never appear on the page.
Perfect post, I totally agree with the Meta Description concept that you have shared with us and really appreciate the way you explain. I myself have seen and tested few of my internal pages where I just placed bla bla in meta description while none of the keywords there in the page’s content. After few days that junk page was crawled but it was not ranking while total SERP was 2.