I’ve always had a gut feeling that identifiers help in adding authority to search results. If a website about a book has an ISBN on it, chances are the ISBN is referring to the actual book. This is true of dynamically-generated product pages at online bookstores – I know from my experience at B&N. And it seems self-evident.
The mystery is that Google is a “black box” – what goes in is not always what comes out, and of course much of their algorithm is proprietary. But there are some things we know, some premises we can make, based on what Google has said themselves, paired with common sense:
- Google prioritizes data it finds valuable
- Valuable data is both unique and authoritative
- Therefore data that’s associated with identifiers (which are by design unique and authoritative) has a better shot at being prioritized higher…than data that doesn’t.
This is not the stuff that’s in the <meta> tags. That stuff has been polluted by spammers and Google barely even looks at it.
It’s…microdata. Well, of a sort – microdata is actually a markup format that describes elements on a website. Elements such as
Kind of a random assortment – but those are the elements that Google is picking up in its effort to create “rich snippets” – those new data elements you see to the right of the search page when you’re looking for something.
Other formats, besides microdata, that help Google accomplish the same thing are RDFa and microformats. But Google prefers microdata, and so that’s what I’m focusing on.
These formats describe the tags that Google will pick up. But what goes IN the tags?
It’s a very vague-sounding word, but for our purposes, an ontology is just a vocabulary that everybody agrees on. Google is using the ontology created by Schema.org. Therefore, if your website uses the same ontology, chances are that “rich snippets” will result.
The important thing from my business’s perspective is that the ontologies include identifiers. Some identifiers that Google picks up are ISBN, UPC and ASIN. Web pages that are about books, with ISBNs in the <title> field, will be viewed by Google as reliable and therefore will be weighted more heavily in search.
When I say “picks up”, I am not referring to page rank. The algorithm for page rank is constantly shifting. But these identifiers do contribute to the formation of “rich snippets”, which in turn call more attention to the web page. Over time, the more attention the page gets, the more hits and links it will get. And those things determine page ranking.
So it’s not a direct relationship between identifiers and search – it’s an organic one. But without identifiers – without that assertion of uniqueness and authority – we’re not even on the road to that organic relationship. We’re condemned to the chaos of keywords.
Many, many thanks to Gary Price, whose conversation with me about these issues clarified my thinking immensely.