NEW DELHI, INDIA: It has been a while since Google had changed its search algorithm as an overhaul; last time was 2001. If you imagine Google Search is a car, then this is an engine replacement. Previously, there have been some transmission and gear box changes i.e. Caffeine (Jun 2010), Panda (Feb 2011), and Penguin (Apr 2012).
But this time, Hummingbird implementation was announced on Google’s 15th anniversary (Sept 26, 2013).
Google has been trying to deliver the best SemanticÂ search (i.e. not just throwing up index pages related to the keywords used, but providing relevant contextual results). Hence, Knowledge Graph matrix was extensively utilized in this deployment. To become truly Semantic and beat the likes of Siri (i.e., Apple OS voice indexing protocol too), Google Voice Search is using the android platform to harvest data that provides contextual meaning to words and phrases.
Example: If I search for “CMO, ABC Company”, I’m probably looking for the Chief Marketer of ABC Company. The old school Boolean search might throw up “No Results Found” or “Contact Us page of ABC Company’s website”; whereas, the Semantic search would list the LinkedIn or name/contact of “SVP/VP Marketing of ABC Company”.
From the data source perspective, all Google applications and platforms are gathering index shards to fuel Maps, Glass, G+, Apps (Free+Enterprise), Enterprise Search, YouTube, Picasa, etc. – all these sources are gathering derivative meanings of words and tagging them to images, documents, and sounds bytes.
They are simultaneously cross mapped to human sentiments and cognitive computing protocols. If you remember IBM Watson in Jeopardy, one didn’t have to explicitly say the “reference meanings” of pronouns, the computer was able to derive the contextual meaning of who was referred to as “he/she” – the same is the goal of Semantic search.
How it affects Search and Content Marketing:
1. Removing Redundancy: If you pitch multiple collateral/assets for the same subject/topic, low-influencing digital assets might be filtered out
2. Increased Metadata Dependency: Whether it is a PDF, a webpage, an image, a flash or RSS, MRSS, Social data … if you have not optimized properly with adequate tagging, anchoring, bookmarking, and meta description, your collateral(s) will slowly phase out from top results
3. Author Rank: Hummingbird will give a lot of weight to the author rank of the content. Your content might be pure gold, but if it’s published by a ‘social/web nobody’ it might be buried under the “in-circle publishers'” posts
4. Content Syndication: If the size (bigger portals) become irrelevant, you will have to fall back on the potency of your content … small time bloggers and user-generated content sites will become more important
5. Increase Google Advertising: When publishing houses become less relevant, marketers would have to spend a bigger chunk of their advertisement budget in the Google products (Not just Adwords, think platform based i.e. mobile and App suites)
6. Apps Marketplace: The API mentality is still in effect where Enterprise Search providers (E.g. BluePoint, Autonomy, Persistent, HP TRIM, IBM WebSphere, FileNet, etc.) already have Hummingbird integration apps available in Google Apps Marketplace. As a marketer, think about BigData and Search integration in your Enterprise CMS environment … a lot of Cross Sales and UpSales is possible in global multinational corporate environment
To conclude, the new Hummingbird algorithm is great for end-users, but we (marketers) need to step up our content development, optimization, distribution network, and reach strategy because, SEO as we used to know since 1994, will soon fade out.
The author heads Communications and Digital Media for a global strategic business unit of Tata Consultancy Services.