The Future of Voice Search

When you’ve been married to someone for 20 years, you tend to pick up each other’s mannerisms and phrases in a way that outsiders might not. Here’s an example:

Dylan: I stopped by OshKosh B’Gosh today.

Nina: Did you find anything? Last time I went in, the inventory seemed a little light.

My husband is neither a toddler nor a farmer, so why would he ever buy anything at OshKosh B’Gosh?

I was able to respond in step with him because decades of knowing Dylan and his sense of humor allowed me to just know that he was talking about the clothing store Askov Finlayson, located in Minneapolis’ North Loop neighborhood. If he had a phone (really, he doesn’t!), would his phone’s voice search understand what he meant if he asked, “OK Google, what are OshKosh B’Gosh store hours?” Most likely, Google would first return the two nearest kids’ clothing stores in the suburbs.

But sometime very soon, search — both voice and standard — will continue to decipher our personalized mannerisms and expressions. The Internet of Things is here, predictive search is continuing to improve, and we’re moving to a world of voice-based search.

Here’s where we stand right now:

1. Widespread adoption, starting with teens: People under 18 may not use voicemail, but they’re certainly more open to voice search. Late last year, Google reported that 50% of teens and 41% of adults use Google voice search more than once a day. We’re expecting this to grow — and fast.

2. Natural Language Processing / Machine Learning: Type a few letters and your auto-fill will predict what the full word is, and also what the next word may be. Google is increasingly testing predictive search, which understands the subject of a query related to a previous search (I wrote on this topic last year). And search engines almost always know all the synonyms and other terms used for a topic, reducing the need to slavishly use them in content copy.

3. Personal assistants everywhere: Recently Facebook announced M and Amazon launched Echo, competitive products with Siri, Cortana, and other voice search services, with an extra feature: Facebook and Amazon assert that their products will actually be able to perform tasks in addition to finding information –such as ordering dinner, buying movie tickets, or creating to-do lists.

4. Voice-powered couch-surfing: Voice search isn’t only for finding “restaurants near me.” Android TVs have been on the market for just over a year. Roku, Comcast, and Samsung have all followed suit in developing their voice-controlled devices. Right now you can say, “Play me last week’s Game of Thrones,” but soon you’ll be watching TV ads and saying, “Look up more information about that car.” (You may feel like a bit of a dork talking to your TV, but your kids certainly won’t.)

5. Even more competitors in the voice search space:  SoundHound, an iOS app that could compete with Siri, is in alpha testing, and Yahoo is developing a personal assistant of its own…


So how can your business optimize for predictive and voice search, and telepathic understanding?

1. Know your customer.  Really dive into who your customers are, and what they are looking for. Keyword research, speaking with front-line customer service and sales, and analyzing  website engagement metrics is a great way to start. Then your analogy would be knowing that when people want a “toy hauler,” they could buy one of your SURVs.

2. Ensure that your website is well organized, your social channels are optimized, and information customers want is easy to find.  If a voice-based search engine is looking up flower delivery companies and purchasing flowers on your behalf — as Facebook’s M is promising to do — you want to be sure that a robot can access your website, your Google business page, your maps and product pages and that a human would want to use it.

3. Quick Answers.  Quick answers are a fast-adopting result on Google for many common questions, and these will become even more important. Ensuring you know applicable questions and use the right mark-up for them will help you both show up in the quick answers and be regularly pulled for voice search “how to” answers.

4. Develop quality content that people want to use and trust.  All of the major search engines, along with networks like Facebook, look at user experience — how people use your content — to determine how often it should show up in a search result, whether activated by voice or by text. Developing good content that offers unique value to customers will help grow your customer base which, in turn, will help the algorithms trust your content over others’.

5. Think about intent. The exact phrasing of a keyword isn’t as important as what goes around that keyword. I’ve written about Natural Language Processing before, but a user finding what she’s looking for is much more important than the presence of an exact keyword match.

6. Be aware of what’s on the horizon. Who are the new competitors in voice search technology? Who will be the first to develop the best personal assistant? Will a newcomer like Hound beat out the old dogs like Google and Apple?

Things are happening quickly — and right now the assistant market is still up for grabs and expanding quickly. I don’t think that our computers will be telling us “I’m sorry Dave, I’m afraid I can’t do that” any time soon — but we’re reaching the point where the intricacies of human communication, even the words and phrases that previously only your loved ones would have been able to understand, may be decoded by your devices.




SUITE 2000

PHONE: 612.392.2427