It’s tempting to think that your boring old metadata doesn’t matter any more. Except it does. It matters a lot.
For the proof, let’s take a look at just one tiny part of the meta universe – your page description.
Meta descriptions – a quick recap
Users see the meta description under your page title on a search engine result page. It’s there to expand on the 66-or-so characters in your page title, giving users a little more detail about what they’ll get when they click – 140 characters of detail, to be precise. Good descriptions should be helpful, clear and – you may want to sit down for this – descriptive.
When Good Meta Turns Bad
If you’ve spent any time optimising a reasonably well-established site, you’ve probably noticed that Google often ditches your carefully written page descriptions in favour of something more to its own liking. Why on earth would it do that?
Because Google thinks it can do better.
Meta source #1 – Google happily uses your page description
Let’s say a user searches Google for ‘public copywriting course’. They’ll see Blackad's page title and description as the first natural result, just after these three ads:
Happy days – the page description exactly matches the one we wrote when we published the page.
Let’s move on.
Meta source #2 – Google uses a sample of your page content
OK – so this time, our user has searched for something a little more obscure: ‘copywriting course irrelevant filler’. They’ll see this:
Again, we're the first natural result, appearing just below those ads.
But what the hell's happening with that page description? Google has taken a look at the search query, and noticed that the page description supplied by us doesn’t contain the words ‘irrelevant filler’. Ever the busybody, Google has constructed its very own page description, by taking short samples from our page copy.
Weird? Not at all. Google does this all the time, and it’s usually a good sign: Google has found the answer to the user's query in the page copy. However, beware: generally, you want to avoid Google doing this for the top search queries for this page. And the way to do that? Understand the queries which create the most organic search traffic to that page, and then tweak your page description to suit.
Or to put it another way, try to stay in control of what users see on the search engine results pages. That means writing unique page descriptions for every page on your site.
Meta source #3 – Google uses content from DMOZ
This is where it can get weird.
Sometimes, Google doesn’t want to use the page description you wrote. Or, indeed, take samples from your page copy.
Why have you been shunned? It’s usually because Google doesn’t think much of your page descriptions or page copy. (There are other reasons, including missing meta description.)
So what does Google show when it doesn’t display your page description or extracts from your page copy? It looks for another trusted source of data: the rather stately Open Directory Project – DMOZ.org
Their reasoning is good: to get your site listed on DMOZ, you send in a listing request, and (eventually), a real human being will review your request. If you’re successful, that very same human will write a short description of your site – this becomes your DMOZ entry. And it’s this entry which Google sometimes uses instead of your own page description.
But there’s a problem. DMOZ entries are often pretty dull. And because they’re edited by humans, they’re always horribly out of date. QVC have this problem with their UK website:
Happily, QVC's is the first natural result when we search for 'qvc uk'. But look at that page description: Telewest and NTL haven’t been around since the mid 1990s. How odd.
Odder still, the fact that ‘Telewest’ doesn’t appear anywhere in the page source (trust me; I looked).
Instead, Google has taken the description from QVC’s ancient DMOZ entry:
How to stop Google using DMOZ data
It's easy to get Google to ignore DMOZ – just add a meta tag to your pages:
<meta name="googlebot" content="NOODP">
But Google isn't the only search engine in town. You can ban all of them from using DMOZ to describe your site. Just use this tag:
<meta name="robots" content="NOODP">
Google's guide to reviewing your page titles and snippets tells you everything else you'll need to know.