Voice and apps, the future of mobile search

Digital Humans / May 2016

Joe Ford

Joe Ford

Need someone to manage a URL migration for a website with millions of pages? Joe's your man. Outside of improving search performance he likes to fine tune his fleet of vehicles.

Tomorrow’s technology, yesterday: the rapid progress of search

There are several aspects of 24th century Star Trek technology that already appear quaintly antique to us here in the 21st, one of which is the method by which the crew stores and reviews data.

Super-android, Commander Data, whose total linear computational speed is 60 trillion operations per second, today appears cerebrally glacial compared to supercomputers that undertake quadrillions of operations p/s. That’s a lot of information processing. This is probably why Data sometimes has to go on the Enterprise’s amusingly anachronistic equivalent of the internet to find stuff out. The enterprise crew also carries information around in stacks of tablets that are quite a lot chunkier than today’s iPads.

Prior to the advent of the internet, the science writers of the late 1980s simply did not predict the speed with which we would overtake Star Trek information technology within only a couple of decades. Today we have at our fingertips more information than ever before, which we retrieve within a fraction of a second. Any longer than that and we’ve lost interest. Furthermore, we expect our searches to bring up information that’s pretty close to what we want. We refuse to spend time ‘hunting around’. We want our tech to intuit our needs and fill in the gaps in our language. Why bother typing an entire command when the core grammatical elements will do? But there is something particularly thrilling each time a little more of our technology shifts further into what had previously been the domain of science fiction as this exchange illustrates:

“Computer.”
“Prrrrb” [hail beep 1]
“Locate the whereabouts of Commander Data.”
“Commander Data is in Ten-Forward. Prrrb.”

There’s something about the voice command that makes one feel firmly part of the future. Hands-free technology is increasingly desirable as we seek to undertake multiple tasks in busy workplaces, and often it’s far easier and quicker to ask a question of our mobile (or desktop) verbally than it is to interrupt our activity and type it out.

What does changing modes mean for SEO?

But there’s more to voice search than the mere novelty of convenience. We talk very differently to how we type: we speak faster; we have accents; we use dialect terms, idiomatic phrases and non-standard syntax; we employ prosodic features like pitch, stress, intonation and volume; we often stumble or mispronounce words and we abbreviate words or leave them out. We also form questions in a variety of different ways:

“What’s the weather going to be like today?”
“Is it going to rain today?”
“Will I need an umbrella later?”

The intelligence of the engine searching for the answer must allow for all of these questions if it is to satisfy the user. There is also the fact that of all searches 10-15% have never been constructed in that way before. That is, people are inputting long-tail search terms that are unique in their lexical arrangement, therefore the engine must extrapolate from the semantic units to come up with the goods.

Typing enquiries necessarily slows us down and there is a moment to consider the nature of what we type, to amend and to clarify, but with voice commands grammatical constructions are spontaneous and the mode allows for more variation. As engines become more agile in anticipating collocations and drawing upon semantic fields and the digital information known about the user through the mobile, so will the accuracy of that search term. The next question might be: how do companies make sure they rank for voice searches?

To boldly go where no search has gone before

Let’s head back to the 24th century for a moment and consider what happens when Picard asks the computer for data verbally: reams of text and diagrams interchange speedily on-screen, seemingly with no consideration for relevance. If this method of ranking search results were the standard today people would soon revert to books for information.

Today, when we request a search result – increasingly using mobiles – we have a very defined list of appropriately ranked information that ideally (depending upon the search terms we’ve used) requires minimal sifting. Users expect to have their queries answered in an instant and by the first few results of page 1. Amidst that information Google supplies not only data from web pages, but also from apps. Google directs the user to the most relevant source of information irrespective of the form in which that information exists. In the case of the above search, Google might direct the user to the BBC weather app result for the location of the user.

From a commercial point of view this is very interesting. If a user were to search for things to do with their kids this weekend, then the result that best caters for their query might well be activities with an adventure company 10 minutes’ drive from home, rather than simply coming up with a variety of only partially relevant options, such as the same activities in the next county. Better still if the user can be served a result linking to a voucher for this adventure company directly within a pre-installed app.

Putting customers at the centre of search

Change is one certainty and the other certainty is (as proven with Star Trek’s technical aspirations) that change takes place in the digital world at a fair lick. That’s why we place so much emphasis on keeping users at the centre of the picture. When users change or develop their searching habits, Google is keen to help them locate the desired information. To ensure your content is still performing in the search results it must also adapt and align itself with this changing nature of search.

Latest news