As with Google's Knowledge Graph, Bing's Snapshot (or "Snapshots" – the label is awkward one) is entity-focused, and the update extends the number of entities being shown, improves and extends the relationships displayed between entities, and displays Bing's prowess with entity disambiguation.
While this is resulting in more detailed and better-connected results being shown for all named entities, the biggest impact is certainly in the display of personal named entities (you know, "people") in Snapshot verticals. And in this they're arguably now doing a much better job than Google is with their Knowledge Graph (and so my reference to a "personal offensive" in the title). With this latest update Bing is more than ever taking advantage of its social partnerships, and is in general exploiting the availability of social information for (living) people.
This is an interesting – and I think substantial – challenge to the Knowledge Graph, which draws its information from a more limited set of ("trusted") sources, and so has a much higher "fame" threshold that a personal entity has to overcome before it can appear in this type of vertical. It may turn out that the Snapshot barrier to entry is as low as possessing a LinkedIn profile.
Bing appears to have effectively removed this barrier by doing a really good job at disambiguating personal entities: if they're displaying information from the right Twitter, LinkedIn, Facebook and Klout accounts, then (regardless of the veracity of the information) they're linking to "trusted" social sources insofar as they're accounts that are owned by the same person.
- From a personal branding perspective, LinkedIn is more important than ever. Aside from the large amount of information that's pulled from LinkedIn into the Snapshot vertical for a personal entity, it also seems to be the location from which the entity's photograph is pulled: no LinkedIn photo, no Bing Snapshot photo.
- Marketers have long pushed back against Klout, arguing (fairly persuasively) that Klout is a pretty blunt measure of social influence. That is to say, the message has ended up being "your Klout score doesn't matter." Guess what? Your Klout score matters – maybe not to you, maybe not to those "in the know," but from a raw optics point of view it's now closely associated with your Snapshot in Bing.
These takeaways are only "practical," of course, insofar as it impacts those who see such results in Bing: currently about 17% of searchers in the United States, and a negligible number elsewhere in the world.
This limited reach may not be the case for long, however, depending on whether or not Google responds in kind. This represents – for the first time, I think – a real challenge to Google's Knowledge Graph results, and I can't see Google sitting on their hands while Bing is now producing better semantically-fueled and interlinked results for non-celebrity personal entities than them.
And Bing is seemingly doing at least as good a job as Google for "famous" people too (for Google and Bing the working definition of "famous" or "celebrity" seems to be "a person with an established page about them on Wikipedia").
Bing might arguably be returning more useful Snapshot results than the equivalent Knowledge Graph results because of the additional social properties (Twitter, LinkedIn, Klout) that are displayed for living celebrities in Snapshot, but not in the Knowledge Graph.
Again, I can't see Google taking this lying down. Bing's socially-connected search results were at one time derided as being not particularly appetizing icing on the search results cake ("I don't care what movies my friends like"). Now the labors of socially integrating entities are beginning to bear fruit, and Google is starting to appear woefully behind in this realm. Google is certainly promoting and expanding Google+ with nothing less than frenetic energy, but even though this isn't a walled garden in the Facebook sense, it is one garden: Bing is now foraging far further afield for social produce to include in its organic basket.
All of this is made possible by Bing's Satori graph-based repository that uses RDF and SPARQL, and has been designed to handle a huge number of triples. Having (at least until now) flown lower under the radar than Google's Knowledge Graph technologies, not a ton has been published on Satori, but here's a couple of articles with interesting insights.
- How Google and Microsoft taught search to “understand” the Web
Sean Gallagher, Arts Technica, 6 June 2012
- Bing Now Knows Much More about People and Places Thanks to LinkedIn and its Satori Entity Engine
Frederic Lardinois, TechCrunch, 21 March 2013
- Knowledge Graph, Satori, and Unicorn
Laurie Sullivan, 21 March 2013
[Update, 24 March 2013: Added the last link above, which also put me onto more Satori information from the Microsoft Trinity project website, which I've summarized here.]
Completely independent of search market share wars, it's really great to see Bing putting competitive pressure on Google through improvements to their products rather than shrill and aggressive marketing campaigns. Search engine users will ultimately be the beneficiaries of more competition in the semantic search results space.