Over the past few years, Google has been steadily making advancements to its industry-leading search product. As standalone updates, none are revolutionary, but in aggregate, they form a much larger picture of the type of engine that Google has set out to build.
Google’s latest enhancement – Conversational Search – enables users on a Chrome browser, Android 4.0+ or IOS Google App to have a conversation with Google. The concept is simple: Ask Google a question and it will answer. Ask it another related question, and it will answer again, and so on and so forth.
Conversational Search: Apple vs. Google
Conversational search isn’t an entirely new concept. Those of us with iPhones will quickly draw the connection from Google’s Conversational Search to Apple’s Siri, a voice-activated personal assistant that is also capable of holding “conversations” with its users. The Siri feature is far from perfect, but true to character, Apple succeeded at setting a trend when it launched the feature along with the iPhone 4.
But while Apple was a first-mover, it’s not as well-positioned as Google to make the most of conversational search. Google’s gigantic catalog of information is much more likely to deliver on the concept, as its wealth of structured data can better support quality conversations or at least an ability to answer direct questions. Google already has the infrastructure in place to leverage information and string it together conversationally.
Below are just a few key milestones supporting Google’s “personal assistant” technology, enabling future conversations with the world’s largest search engine.
- June 2011 – Voice-to-text Search “Google Voice Search”
- June 2011 – Social Search “Google+”
- May 2012 – Semantic Search “Knowledge Graph”
- July 2012 – Predictive Search “Google Now”
- May 2013 – Natural language Processing “Conversational Search”
Why Conversational Search Matters for Marketers
Voice-to-text use among searchers can be expected to grow following this announcement, which will lead to a shift in the types of queries submitted to Google. Currently, about10 percent of searches on the iPhone, and 25 percent on Android devices, are triggered by voice. As conversational technologies improve, consumer search behavior will change and our voiced queries will produce different outputs, such as longer tail and/or more fragmented queries.
Additionally, conversational search will underscore the importance of structured data on a marketer’s website. Schema.org and micro-formats should be adopted in website page development. Given that the new feature will likely encourage people to submit queries more conversationally – as a thread of questions – the more details available to Google, the better.
Currently, conversational search is limited to fact-based queries (e.g. “How old is Barack Obama?”), but it’s reasonable to assume that someday this could extend even further. For example, consider someone shopping for a dress via conversational search. She would likely use a string of queries to narrow her search to a specific color, size or brand. Marketers that incorporate rich mark-up data will better support conversational search and other advances in Google, allowing brands to surface more content as users filter down their path of discovery.
Finally, marketers should be cognizant of the potential disruption to analytics that might result from an increased usage of conversational search. Continued challenges with “dark Google” (Google Secure Search) will only compound for signed-in mobile users – and for those not signed in, know that the keyword logged in the referrer string might only be a partial term of a larger keyword query.
At this time, it remains unclear how Google will pass keyword strings of queries via conversational search versus traditional search queries. In the dress example above, will all queries in the thread be seen by analytics tools, or just the final query in the conversation? There is a chance that web tracking and analytics tools will only receive the last spoken keyword versus the full context.
As Search Engine Land’s Danny Sullivan noted, “search veterans” are hard to impress, and once you use this new functionality, you feel the potential power it offers into the future. Google Now is leading the “predictive search” revolution, so layering on a conversational functionality is a great next step for Google.
This more powerful version of the already-familiar Siri is now being placed front and center on the world’s most popular search engine, priming it to considerably disrupt the search landscape in the near future. A significant share of people will opt to “talk” to Google versus typing queries from a keyboard, even if the feature is only used under certain conditions (e.g. not everyone will want to voice their private queries out loud). Actual user behavior pending, conversational search clearly offers a faster and more natural way of filtering and discovering information online.
Cover photo via Facebook