Marketing AI Institute | Blog

Intelligent Assistants Are Key to Making the Most of Your Data

Written by Duane Forrester | Jun 30, 2020 4:26:00 AM

Whether you’re a complete convert or remain a skeptic, today’s intelligent assistants are here to stay.

Google Assistant, Siri, Cortana and Alexa are the standard default helpers in our everyday devices — designed to help us find, execute, discover, shop, and solve elements of our everyday lives.

And oh boy, do we want to use them! eMarketer recently released a report that noted 35.4% of the US population are voice assistant users. This is a 5.3% increase over 2019—and the number is expected to keep increasing into 2021.

The emerging bottom line is clear: More people than ever are using voice assistants as part of the process to accomplish tasks. From shopping, to sending money, to booking appointments, and more, these systems are becoming a deeper part of our daily lives.

How to Achieve Success In Voice Search

To be successful in voice search, you need to invest in several areas—some technical, some less technical—designed to pass tests run by humans. These humans would be the Quality Raters at Google, for example. Their job is to review select results to ensure quality guidelines are met. This not only helps with overall quality in the search results, but also helps train the machine learning systems in the background.

In the past, Quality Rater guidelines strictly focused on organic search have surfaced, leading SEOs and digital marketers to a deeper understanding of what it took to rank well in organic search. Today, we have a set of guidelines for Quality Raters at Google which are focused on voice search.

Let’s take a quick look at three major areas of focus from those guidelines, as well as three broader best practices that support efforts to rank well in voice search instances.

Google's voice Quality Rater's guidelines point to three principles that are used when reviewing voice search results. You need to understand them, and learn how to serve up high-quality answers based on these criteria.

  • Length: The length of an answer should be relative to the complexity of the question. A complex question is rarely answered with a simple yes/no, and the engines are looking to understand if you are taking a shortcut, or if you are providing a useful service to a searcher.
  • Formulation: Here we’re focused on grammatical accuracy and response formulation. In other words, would a native speaker respond in this manner? Included are any attribution elements—did the answer need an attribution, was it properly called out and understandable?

Some results are also judged against:

  • Elocution: For this element, the rater is looking to hear how well the response is spoken, as a phrase overall, and each individual word. They want to hear that the response sounds natural, has correct intonation and inflection, and that the speed of speech is reasonable and appropriate.

The next three elements should be very familiar to folks who optimize for search. Your site needs to be:

  • Mobile-Friendly: This is not optional anymore, especially as Google moves to their announced mobile-first strategy. Essentially, if you don’t perform well in a mobile environment, your ranking will suffer overall. Desktop will be influenced by mobile performance.
  • Secure: This has been a consistent drumbeat from Google for several years. They believe an HTTPS environment is better for consumers, and it would advantageous to align here. You don’t want to be the last of your industry to remain non-secure, as you’ll be squeezed out of results over time. There isn’t a penalty here, but if other sites can answer the query just as well, and they are secure, they’ll outrank you.
  • Marked Up: Marking up your content with Schema.org elements (or elements from another source) is becoming increasingly important. It’s a signal that has growing trust from the engines. Given the consumer demand for rich experiences (this marked up content powers those rich experiences) its influence will continue to grow. Skip this and you’ll fall behind.

Once you’ve mastered those elements, you’ll be at the front of the class in terms of being spoken out loud by digital assistants. You’ll rank better and see more traffic than before, and gain voice-enabled engagement.

Don't Exclude Digital Assistants From Your Strategy

Do not make the mistake of thinking you can skip this evolution. Today’s digital assistants, while largely robotic in nature, are rapidly growing to become tomorrow’s digital agents. They are services that can act on our behalf without hands-on guidance for each step.

Tomorrow’s digital agents will be the personal gatekeepers for accessing consumers. If you can’t prove value to the search engines, you’ll never prove value to a digital agent, and you won’t reach consumers when and where it’s most appropriate and beneficial.

This digital agent shift is still a couple years away, but we’re already firmly on the path there, and only comparatively modest growth in machine learning is needed to bridge this gap. You can be sure the major players are working hard to get there, and secure the loyalty of consumers who prefer their version of a digital agent.

Want to learn more? I break down even more tips and insights on the future of search and how your business can harness this technology to reach searching consumers this ebook from Yext.