How to use machine learning to improve UX and on-page SEO | Articles | Strategy

With major algorithm updates being rolled
out every other month, SEO has become a game of keeping up and anticipating
changes.

The latest and the most significant
disruption for marketers has been voice search. Now that Alexa, Google
Assistant and Siri are listening to user questions and learning how to respond
to more complex commands, businesses should prepare for this new type of
no-screen selling experience.

Just how big is the voice search
opportunity? Per
Google,
72% of voice speakers owners are using their devices as part of their daily
routine. What’s even more promising is that a lot of voice users are very
interested in interacting with brands:

Survey question: What
voice-activated speaker owners would like to receive from brands?

Source:
Think with Google

Commuter commerce is another multibillion
dollar opportunity for brands. According to the
Digital Drive Report 2019,
53% of drivers interact with voice assistants on the go. More than 30% of
respondents also regularly use voice technology to place a drive-through order
for food, coffee or groceries.

The challenge for businesses, however, is
that when voice searches are conducted through AI assistants, only the first
several search results stand a chance to be rendered to the user. Getting on
that first page of Google SERPs is critical. So how do you win over AI? Your
best bet is to rely on the same technologies the smart assistants are using
–
machine learning (ML) and deep learning. Below you will find the four use cases of ML and AI toward
on-page SEO and UX.

1. Switch to predictive analytics for
richer customer insights

Traditionally, marketers have used customer
personas to determine how they will deliver value to their target audience,
based on the descriptive online analytics solutions (e.g., Google Analytics)
and via offline methods (e.g., focus groups, interviews).

However, customer journeys today have
become
more complex and omnichannel. One shopper can experience multiple touchpoints
with a brand before converting – all taking place through different devices and
channels. To capture and analyze those behaviors, businesses need more advanced
tools.

Predictive analytics solutions allow you to
capture a wider range of customer activities online – what they have looked at
and purchased, where they hang out online, and predictions of what they may
seek going forward. All of this data can be gathered, sorted and spat out based
on specific queries and transformed into applicable business insights.
Predictive analytics can help you determine:

• The products and services that should be pitched to niche
  audience segments.
• The types of content that marketers should be creating to
  engage with buyers at different stages of their purchase journey.
• The keyword opportunities worth pursuing to attract segment X.
• Improvements that can be made to the overall user experience to
  drive further conversions.

2. Prepare your content for voice search

No longer is a searcher simply typing in “light
fixtures”. They are now speaking to their smartphones or Alexa units with phrases
such as: “Alexa, find me stores that sell kitchen light fixtures in
Dallas, Texas”. Alexa will then conduct a search using the terms “kitchen
light fixtures” and “Dallas, Texas”.

Companies are now faced with the new task
of optimizing for those natural-language queries with little-to-no guidance or
tools from Google and the like. So far, several things have proven to
help business rank well in search results for spoken questions:

• The language of voice search is critical to the content that
  you produce. AI platforms respond in human terms and they use the
  vocabulary in the search terms to find page texts that have the best
  match. Marketers will have to research and anticipate the words that a
  searcher will use to find them. Specifically, focus on answering the who
  what, where, when and why types of questions associated with your
  industry.
• A site must be fully mobile-friendly, because more voice
  searches will be conducted on smartphones. If a company relies on local
  traffic (e.g., a restaurant), Google My Business profile optimization will
  play a key role.
• Craft your business FAQ pages carefully to answer those “W”
  questions mentioned above. They should include those long-tail keywords
  that are used in voice searches. All other landing pages must do likewise.

3. Invest in personalization/recommendation
engines

If you wonder
how Amazon recommends additional products for you to purchase, it is the
result of analyzing your individual purchasing behaviors, comparing them with
other consumers who have purchased those same things and then generating
suggestions for you personally. All of this is the result of AI and ML.

Building or adopting a recommender engine
may seem like a costly endeavor for brands. However, the ROI of such systems
are rather lucrative:

• 48% of consumers spend more with an e-commerce company offering a
  personalized shopping experience.
• Product recommendation systems deliver a 23% lift in conversions rates for web products on average.
• 71% of consumers state that personalization would influence their
  decision to interact with the brand’s emails.

You can achieve similar results by personalizing
messages and content even in real time, providing a user/customer with the
personalized experience based on their needs. AI can also be used to determine
what action a business should take next to move that target further through the
buyer’s journey.

4. Deliver better customer support
without increasing your team size

Nothing can kill a relationship with a
potential or current customer more than a poor customer support experience.

Traditionally, businesses have used call
centers, live chats and email to answer questions and resolve issues. These are
still good venues, but AI and ML bring great new technologies to this process
and can result in great savings at the same time.

Gartner predicts that
by 2020, 85% of customer support relationships will occur with no human interactions. Enter the world of chatbots. In fact, they are already here
in so many ways – everything from Poncho the weather forecaster, who learns,
adapts and converses with and entertains his users, to Tacobot, who takes
orders remotely and suggests more menu items. The combinations of AI, ML and
natural language processing (NLP) allows these bots to continue to learn and
improve interactions with and support for users.

As AI and ML continue evolving, there are almost
unlimited opportunities for companies to retrieve and use data and analytics to
gain insights on their customers, to craft amazing customer interactions
through meaningful messages, and to enhance customer engagement and
experiences. In terms of SEO, new search activities of consumers will
drastically change the landscape of getting “found” and achieving top
rankings.