A machine will decide whether your business is seen by others. This is the reality we live with today, and its application is expanding.
Never before in history have we produced so much data. As a period in time, right now is a watershed moment simply due to the volume of data we produce and consume. All that data is tracked, watched, analyzed and any patterns that can be found are extracted. This is the very core of machine learning—sorting through data to reliably identify patterns and predict which pattern will occur next.
And machine learning is what we really mean when we talk about AI. It’s the unsung hero in the background, collecting the data, parsing the data, analyzing the data, and coming up with answers.
If you’re a search engine, or a service that tries to answer questions for consumers, your understanding of what answer will please the next consumer often grows from input on what the last group of consumers thought about the answer you provided.
This is the basis for today’s world of search—a world where daily tests are run to validate every theory or enhancement. As the group of people exposed to a test scenario expands, the confidence grows on whether the change was positive, negative, or neutral. Those results are tracked, the patterns analyzed, and decisions made on what to do next. Today, more often, those next steps are decided by machine learning algorithms and systems. This means the cycles can move much faster, which in turn means results can be moved to positive outcomes much faster.
While it’s fascinating to dive deep into machine learning and its possibilities, we can’t forget the business impact.
How is Machine Learning Applied to the Ranking Algorithms in Use Today?
Crawling the web is a known constant. Generally, even the newest of SEOs can position a business to be easily crawled by following the standard best practices. This essentially means exposing the business’ data for consumption by a search engine crawler.
The next step in the process of ranking is indexing, which is when content is sorted and stacked in ways it can be easily retrieved on demand. As you can imagine, index sizes at the search engines are massive, so machine learning can help systems intelligently manage all this data. Duplicate suppression, quality testing, refreshing, and more can be triggers—and managed by cues from smart systems.
Then there’s ranking. Machine learning’s influence is heaviest in this area. It makes sense to apply machine learning at this level because this is also where data testing is done with live results. Of the millions of queries run in a given hour, an engine can run hundreds of tests at once, collecting data and making decisions in near real time on things like font colors (and sizes and styles), pixel spacing adjustments, quality changes based on algorithmic changes, and so on. And those little things like pixel spacing and font colors have a big impact on where people click, so it’s all tested.
It's in the area of ranking that AI has the biggest impact on a business because we’ve got another layer to contend with today: the group of digital assistants that inhabit our phones and the smart objects in our homes. Those items are always watching, listening, and collecting data on their users, and then feeding it back into the larger system. And its that data flow, from an individual person into the system, that makes today the most epic moment in history.
Just For You, Dear Consumer
With the convergence of data from broad system learning and our personal data streams, search engines and systems have a clear path, not only to telling a consumer that, yes, a pizzeria is nearby, but to crafting a personalized answer based on past data that allows the system to tell the consumer the best pizzeria for that individual consumer. Some people value price more than star ratings. Others want an authentic style of pizza, while still others make decisions based on take-out or dine-in as preferences.
In this world, businesses not only need to employ common SEO best practices to rank, they also need to review the quality and depth of their content, how well that content works in a mobile environment (because that’s where the majority of internet traffic exists today), and how unique and useful the content is. This means that having a mobile-friendly website is now a requirement.
Since the search engines want businesses to be on secure connections, having a secure (or HTTPS) site is important. It’s not mandatory, but Google’s Chrome browser is now starting to flash warnings to surfers who visit non-secure sites, telling them the site is not secure. This could have an obvious effect on traffic, so better to be secure.
Finally there is the effort to mark up your content. This ensures that discrete elements of your content are wrapped in trusted tags, identifying them as what they say they are. This is technical work that happens at the code level of your website, but it’s important to do. Both Google and Bing support all markup languages, but they both jointly run www.schema.org and suggest this language set for marking up elements. The list of what can be marked up is long. Hundreds of individual items from objects, to content types, people, places, things, times, locations, and so much more.
And its important to mark up your content because that marked up data is relied on by the engines when answering queries. Skip this step, and you look like a poorer result to the system tasked with answering a searcher’s query. Fill in the blanks and you get to participate in the rich mobile experiences we enjoy on our phones today.
AI’s Next Gift to Businesses: A Trust Gate
As AI systems become truly intelligent and capable of making realistic, day-to-day decisions on our behalf, we’ll see today’s digital assistants grow into digital agents, managing our mundane tasks each day.
Part of what they will manage will be your calendar. They’ll be able to block or accept calls and prioritize incoming messages. They will sift through our email, dumping the junk and telling us what needs attention. In short order, these digital agents will be the gatekeepers to our lives. They will function as an active trust gate, letting into our view the things we teach them matter to us, and teaching them to skip the things we value less.
As a business, this means building a solid relationship with customers, so those customers know to tell their digital agents to let you in; that you’re a trusted friend. All of the steps that a business takes today to help current systems identify their usefulness to consumers today will help build trust in the business tomorrow. And that trust will be needed in the future.
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 in this downloadable ebook.
Duane Forrester
Duane Forrester is VP of Industry Insights at Yext, a company pioneering a new category called Digital Knowledge Management, which gives businesses control of all of the public facts that they want consumers to know across the intelligent ecosystem.