HUMMINGBIRD FOCUSES ON THE ‘MEANING’ OF A SEARCH
Hummingbird uses Google’s Knowledge Graph, which was introduced last year. The Knowledge Graph allows the search engine to better understand the relationship between concepts instead of individual keywords. Using this, Hummingbird was designed to focus on the meaning behind the words instead of just the words themselves.For example (to quote Danny Sullivan of SearchEngineLand’s findings): A search for “acid reflux prescription” used to list a lot of drugs (such as this, Google said), which might not be necessarily be the best way to treat the disease. Now, Google says results have information about treatment in general, including whether you even need drugs, such as this as one of the listings.
Crucially, this allows webpages that match the meaning of a query to rank better than a page that matches just a few words.
Stephen Bateman recently posed the question, “Is semantic relevance the new Google linchpin?”, to which we say an emphatic, “Yes!”.Semantic relevance has been a linchpin that Google have been plotting for over a decade. Google’s Amit Singhal first filed a patent called Search queries improved based on query semantic information in March of 2003; development of this technology has taken a long time, and is a very big deal. But beyond a better search experience why are Google doing this now?
SEMANTIC RELEVANCE IS A TROJAN HORSE FOR PREDICTIVE SERVICESThe use of semantic search in mainstream search engines is considered to be the holy grail of search engine technology, yet Google is not content to apply semantic relevance to one vertical. Services such as Google Now are an example of Google increasingly using semantic web data (not just to search) to predict where an individual is going next going next – both on and offline. Google Now consolidates local data (agenda, location) from your smartphone, and combines them with on-demand database content (of selected databases such as airline flight-information) and services such as Google Navigation and Google maps.
All these pieces of information are collected, not only to make results comprehensive, but to actually predict what you might need next.
AN OPENING SALVO TO DO BETTER CONTENT MARKETINGGoogle’s commitment to serving increasingly personalised – and increasingly predictive – content will mean that brands will have to cultivate a mindset where they are producing useful content (often tangential to a specific product) as they seek to engage browsers earlier in the purchase journey. At idio, we often talk about moving clients from ‘retargeting to pre-targeting’. Understanding someone’s personal context through their content consumption is a great way to better target them with content that keeps them engaged and prevents ‘interest abandonment’. As Google get better at predicting (influencing!) what we want, the onus is on brands to create (and curate) enough content that can be relevantly served to meet a particular individual need.
INTEREST DATA FROM CONSUMERS’ CONTENT CONSUMPTION IS COMMERCIALLY VALUABLEAs the semantic web becomes more prominent, brands and publishers will surely consider the commercial value of opening up to the search engines and make their data available (e.g. live train info), with new ways of assigning access or restricting access based on personal subscriptions, traits and preferences.
Furthermore, whilst consumer social, transactional and behavioural data have all historically been interesting data pools for publishers and brands to expose to ad networks, interest data derived from consumer engagement with brand content will become the next goldmine (read more on this in our Guardian article).In light of October’s Hummingbird update, you’re sure to hear a lot of confusing marketing/tech speak about over the next few weeks.
Simply put, Google is trying to understand the human need, and provide that person with what they need. Brands take note.