Structured data opens a ‘can of opportunity’ for paid search

Posted on 15. Mar, 2011 by in Thoughts



Semantic Web Logo Structured data is great! It takes information you and I (humanoids) understand, and makes it machine (Matrix style) readable. “Machines” for the topic of this post, search engines, do a good job understanding content. Take Bing and Google for example, they crawl and web and are able to rank documents in an order based on what they (an algorithm) deem relevant.

Search engines do this well, even though they’re inevitably “blind” as to what content really means. By blind, I mean they don’t know “that” postcode is related to “that” hotel, or “that” price is associated with “that” product. We humanoids have eyeballs and assistive technologies (brains) that make those connections.

Structured data like microformats, RDFa, etc should be thought as the “brains” of search engines, enabling them to make those assistive connections.

If you’ve nursed on the tit that is Google’s Local, Mobile or Product Ad Extensions, you may (or may not) have noticed a particular trend – data feeds are at the foundations.

Data feeds are the same as structured data in the context of this post; they take normalised data, structure it with labels and meta data for search engines to understand.

For example, Local Ad Extensions, you really need, but not limited to, supply a Places feed with all your local business data. From that, Google will extract data from the feed you supply and automatically manage your placement – you just determine the monthly spend.

Google does this by extracting context from your data feed and apply it to user queries. For example, Google will use your hotels postcode spliced with the user’s location to suggest the quickest route to the hotel or to customise the telephone number in the Click-to-Call ad – Google does all this magic automatically for advertisers (we agencies are at the pub when this all goes down).

The process is the same for a lot of other Ad Extensions, Nearby-By Store or Multiple Locations, to name a few.

Well, what happens if you introduce further levels of structured data into paid search? Forget about structured data in the form of “feeds” for now, we’re talking hard-coded XHTML – the good stuff.

Structured data (recap) like RDFa, microformats etc would make any form of content readable by search engines. If search engines understand the content, they can then relate that back to advertisers.

Exhibit A – bigmouthmedia’s address built using a hCard generator.

Visually, you and I don’t see what’s under the hood, no, we just see the address details, but a search engine – they know the postcode EH6 6QH is associated with bigmouthmedia Edinburgh.

Bigmouth hCard

So what does this malarkey have to do with paid search? Well, with a wide array of AdFormats, and more likely to come, could structured data do a lot of the work for advertisers. Let’s take a look at some examples.

  • Click-to-Call – automatically extract the telephone number from the local landing page and Google manages whether click-to-call ads are shown. Naturally, you can turn this off or on.
  • Video – within Google’s Display Network, based on your video RDFa, Google will extract your video, keywords and automatically bid on your behalf.
  • Product – having applied the hProduct microformat, Google will use the price and other facets to enable the advertiser to create dynamically updating AdCopy.
  • Review – promote products when they have a review of 3 or greater based on hReview.
  • Event – a new enhanced AdFormat detailing the time, location and Facebook style R.S.V.P. based on the hCalendar microformat.
  • Social – applying Google’s Social Graph API, which of your followers have shared or even visited the advertiser.

The list is endless (literally), and we’ve only touched on current microformats – more are in proposal stage – take hNews for example.

All the advertiser would need to do is manage the ad spend, potentially which landing pages are promoted and in which manner. Google and its crawlers would do the rest (pub again, anyone?).

AdWords is still daunting for the Average Joe, but if AdWords took a lot of the manual labour and complexity out, then it would be much easier to use. We know Google would love to reach the 1 billion businesses that don’t use AdWords.

The theory of paid search using microformats might be just out of sight at the moment, but with HTML5 nearly ready (<– view the source code) and search engines emphasis on structured data, it is hopeful that someone brainy at Bing and Google have looked at the connection between paid search and structure data.

What do you think; does structured data have a place in a paid search?

Tags: , , ,