Getting Sentimental about  #Sentiment  - do you use it to help disambiguate your entities?

I never thought I'd get into #SEO . As a writer of short stories and poetry, when illness forced me to look for a career in online writing, the last thing I thought I'd be studying was "What makes content rank in the Search Engines?"

The past
When one webmaster plucked me from the obscurity of oDesk (and from earning $10/500 words - seriously - yee-ouch!) and gave me an "SEO Template" to work to, I thought it was the greatest thing ever.

Little did I know then that the linking strategies he gave me to accompany the copy I was writing would later be smashed by Mr Penguin. Like I say, I was new to the game, then.

In retrospect, I'm thankful that all of the articles were published under a generic author name. Yes, even despite ranking 3rd for a 60k/mth search term on the bones of a WordPress template and no off-site SEO to speak of, I'm happy that content doesn't bear my name.

The present
To today: the mix of awesome content, social signals and the semantic web all affect #SERPs  with seemingly greater impact every passing week.

The learning curve of writing to optimise content for organic discoverability, rather than declining in gradient, extends at an almost perpendicular angle, towers into the clouds and disappears into the ether, peaking somewhere around Saturn, I guess.

Today's task was simple enough. And one I've undertaken thousands of times before: compose the social snippets for content I'd written for a client.

No sweat, right? Erm, not today for some weird reason…

…I popped the content into one of the #NLP  processors we use (too idle to find the screenshot from when I composed the content) and - for whatever reason - I wanted to know why the social tags it was suggesting were rendered.

For sure, they made sense when aligned with the article. But the keywords were not the exact phrases used in the copy.  

I started to look at the #taxonomies  and the relationships the subject categories had with the #entities  within the content.

Again, nothing new there. It's something I do daily. But what exactly was "Targeted Sentiment"?

A step towards the future…
Today has been a fantastic day of discovery.

Beyond the tools we use, I ventured into Google Search, happened upon a great SlidesShare from 2009 ( along the way, which pointed me to journey's end, here at +SEO by the Sea.

Once again, +Bill Slawski is the man suggesting the answers - albeit that the below post dates more recently (2011) than the one referenced in SlideShare.

A brief overview of sentiment
Sentiment can be used by the search engines to help support (or provide) reviews that we see in SERPs. Those reviews don't have to be from TripAdvisor (as Bill exemplifies) or official review sites.

If Google finds an entity (corporation, firm, brand, person) referenced in an article and that article uses sentiment to describe that entity (e.g. "Gold Leaf used to be a great smoke" or "West Brom have been awful this season" [tee-hee]), that sentiment is understood by the indexers.

Whether Google chooses to render that sentiment is up to them, of course. And no doubt the way an individual search is worded may return different sentiments (that's my guess - Bill doesn't suggest that, just to be clear).

There are ways today that you can dictate what is placed in those search snippets. Google Local, RDFa vocab and schema mark-up seem the to be those (although they're all still somewhere in the stratosphere on this learning curve pole).

But one thing is clear. If you're writing for your client or the industry they're in, be sure that the spin you're putting on the adjectives to describe them is positive.

You wouldn't want a bad review of your client's industry pulled in from content you'd written, would you?

The best way to ensure you and the algorithms are on the same wavelength is to test your content's readability first and make sure you disambiguate your main entities and keywords.

However, that's a post in its own rite; I've waffled enough already.

As well as thanks to Bill, a big shout out to +John Britsios, without whose input and guidance I'd probably still be copywriting to the Penguin-smashed SEO template from all those years ago.

Now, can someone fetch me a ladder to get to the next level, please?

Exhausted - catch you tomorrow…
Shared publiclyView activity