Semantic Image Search and the New SEO [SLIDESHARE]
Over the past few weeks Google has delivered a classic one-two blow: First search is now delivered over an entirely encrypted connection making for the now infamous “100% not provided keywords” in Google Analytics and second, Matt Cutts announced that the next PageRank (PR) update “may take some time.”
Both of these are related and are, in turn, directly connected to Google’s “going dark” policy where algorithm updates like Hummingbird are announced sometime after they have taken place (if at all).
Google is systematically removing all the tools that traditional search engine optimizers and marketers had at their disposal that allowed them to reverse engineer search and create a set of metrics that could be used to gauge progress in search rankings for their clients. The reason Google is doing that is that metrics of the kind we used to have, created a leaderboard that allowed marketers to do exactly what Google advised them not to do. Namely “chase search engines” instead of delivering value through content.
Hummingbird is clear evidence that Google search is getting smarter, drawing from a much more enriched set of entities in its semantic index and a much deeper Knowledge Graph. An Entity is a concept, divorced from its linguistic counterpart and defined by the relational amount of data that form its attributes, properties and uses. And Entities are now powering Google’s Image Search helping the search engine recognise images in pictures the same way you and I would.
This is important for marketers for many reasons:
- Pictures can now break free of the restrictive and sometimes cumbersome picture description and actually function the way they should. A picture of a couple on a beach, for instance, can now appear in searches that feature, holidays, beaches, romance, relationships, and (perhaps) the particular locale the photo was shot at, if Google recognizes it.
- Photographs become important because they can lead to the surfacing of content. This means that they should ideally be as unique and value-laden as text.
- With improved object recognition in Image Search Google can also begin to better understand text on a photograph paving the way for memes to become true viral marketing aids with a distinct message that can be indexed and served in search.
Google has added one more spice to the mix, by also displaying Google Authorship on Image Search for images that appear on pages where Authorship has been claimed by the page.
The SEO landscape is now changing. Text can no longer be illustrated by bland stock pictures that do little to enhance it and which fail to tell a story. SEOs can no longer rely on checkbox optimization with content created around keywords and link-building as the primary driver for ranking in SERPs.
In the semantic web a company’s or brand’s digital presence is becoming as unique and discrete as it already is in the physical world. And just like in the physical world two businesses cannot occupy the same physical space so in the digital one each business will have to clearly define its identity and its unique selling proposition, just to stand a chance.
For an introduction to Entities and SEO check out my SMX East 2013 presentation:
David Amerland's latest book is "Google Semantic Search: Search Engine Optimization (SEO) Techniques That Gets Your Company More Traffic, Increases Brand Impact and Amplifies Your Online Presence".
He helps multi-national clients and start-ups to organize their SEO and Social Media strategies. He is a business journalist, author and international speaker. He blogs about social media and search ...
Other Posts by David Amerland
Social Media Today