Have you thought about how artificial intelligence (i.e. robots) might impact your business in the coming months / years?
I hope so, as the rise of artificial intelligence is undeniable in the digital marketing world. In just the past few years, artificial intelligence has already started to make us think about what can and should be delegated to automation controlled by AI.
In this article I will explain how online marketing with artificial intelligence will allow you to drive better results through more quantified and automated strategies. More companies than ever — from Google and Facebook to your aunt’s crafting business — are turning to AI to improve their business and marketing tactics and I hope to help you do so as well.
A few years ago, everyone was talking about the rise of big data. Businesses would soon have more information about the industry, their customers, and even employee performance than ever before.
While piles of data are exciting, creating hypotheses and algorithms to sort that data is a laborious and human-dependent job. A savvy mathematician can model one set of data per week, but AI technology can model thousands in a few days.
By now, most industries have a means of collecting customer data and are starting to use AI technology to streamline their processes.
For example, in the health-care field, patients answer a series of questions about their condition, and a robot analyzes their symptoms, medical history, and outside conditions to provide a list of probable diagnoses. It also provides the probability of each diagnosis, outlining steps that the doctor should take to confirm it. Health-care providers plan to spend more than $6 billion on AI development through 2021.
On the tech side, companies like Facebook and Google aren’t much different from health-care robots.
Google’s RankBrain algorithm uses machine learning to determine exactly what search results you want to see. Facebook uses AI to create customer service chat bots in its Messenger platform.
In both cases, AI is cheaper than hiring developers and mathematicians in the long run, and a better user experience increases traffic and thus advertising dollars.
What do customers have to say about businesses’ harnessing their data and using AI to help their experience?
In the sphere of online shopping…
80 percent of customers said they weren’t happy about companies collecting their data but had resigned themselves to it.
In brick-and-mortar locations…
Almost 70 percent of respondents were unaware that they were being tracked but said they wouldn’t change their shopping behavior to avoid being followed.
This means customers have accepted that their data will be collected, caring more about the product and the customer experience elements of the transaction.
By 2020, 85 percent of business transactions will occur without a human, making customer service the final frontier in competitive industries.
With 70 percent of customers willing to pay more for quality customer service and 86 percent willing to leave a company over bad service, businesses will want to invest in AI to make their customer experiences flawless.
For online marketing, the time to invest in AI was yesterday. It’s not too late though.
Industry leaders already use AI to monitor customer behavior on their websites and identify ways to make improvements in the shopping experience. On other marketing channels, brands use AI to test new advertising before a full-scale launch.
AI is the newer, cheaper market research model.
While businesses have used the internet to A/B test creative ideas for years, the process of identifying a successful concept has changed through AI. Historically, marketers would measure clicks and conversions for their display ads and use customer actions to determine success.
With AI, facial micro-expressions tell marketers what they really need to know about customer reactions.
By reading emotions like anger or joy in the human face, computers were able to discover what type of washing detergent scents people prefer. The AI model directly contradicted the survey and focus group information the company had collected but was overwhelmingly correct.
Instead of guessing through flawed customer-response data, AI lets businesses make smart decisions based on fact.
AI means better decision-making to create a better customer experience. This doesn’t mean that market researchers will become obsolete, but they will have better tools to ask the right questions and report back to management. With more reliable knowledge about what a customer really wants, marketers can tailor their message without seeming intrusive or creepy.
Customers only notice AI when it’s done poorly. If done well, the experience between customer and company can be seamless.
In November 2015, news broke that Google had a new tool for sorting through search results. RankBrain uses machine-learning AI, as opposed to algorithms hand-coded by developers, to determine a site’s SERP value.
Most of Google’s algorithm changes come from human hands. Developers create stemming lists and make database changes to constantly improve search results. However, this is still a losing battle.
Google processes 3 billion search results per day, 15 percent of which have never been seen before. There’s no way Google can continue to grow if human hands control the algorithm behind 3 billion searches.
To make RankBrain, humans created a small set of “if–then” rules and then let the computer begin testing and learning from them. The idea behind AI is that the more information the computer encounters, the better it gets at making decisions.
Eventually, the machine creates its own “if–then” rules and is able to make decisions by itself without a human uploading a list of possibilities. Google plans to use this machine learning to tackle the 450 million new searches each day.
The goals behind RankBrain are simple: Harness machine learning to provide better search results for users and reduce the burden of algorithm development on company employees.
While the latter goal won’t affect businesses, the former definitely will.
Of the hundreds of ranking signals that Google listens to, RankBrain skyrocketed to the top three.
Furthermore, Google tested its engineers against RankBrain to look at a few listings and predict which ones would be at the top.
The humans had a 70 percent success rate, but the machine was right 80 percent of the time.
For businesses worried about their SEO, RankBrain isn’t as scary as it seems.
Google’s AI is only being used on complicated long-tail keywords for now, so you don’t have to worry about your core terms taking a hit.
When RankBrain does take over more of the search results, pages will be ranked on a scale of 1 to 10 (with 10 being the best), similar to a Google Adwords Quality Score. Playing with the idea of an “organic quality score,” websites that have high click-through rates, strong conversion rates, and appealing headlines will see the best results.
Google is using RankBrain to better evaluate intent. AI will read convoluted long-tail keywords to understand what the user is looking for and serve results that read between the lines to find an answer.
On the publisher side, RankBrain evaluates the intent of a website’s content and determines whether it’s a good fit from a SERP standpoint. This means publishers who build their content around helping their customers will see the greatest benefits from the RankBrain change.
To prepare for RankBrain, run a health check on your site to make sure the pages load quickly and responsively.
For your content, evaluate each page to make sure it answers all the questions your visitors might have when they land there.
If you run a paid search campaign, use search terms to make sure you’re not driving unqualified traffic that’s increasing your bounce rates, and consider optimizing for a few underperforming keywords.
Google didn’t look to AI to make an impossible-to-beat algorithm update. It created RankBrain to think like a human and serve results that a human would want to see.
If you focus on building a quality website that provides a great experience with useful, engaging content, then you will be on the right path.
If Google uses AI to rank websites, could marketers fight back by creating AI-produced content that robots like?
More developers are creating AI tools that generate content — free of grammatical errors and typos — whose style mimics the publisher’s. Both Karstad University in Sweden and the New York Times have put these robots to the test and found that humans can’t tell the difference between human-generated and robot-generated content.
For the time being, news outlets benefit the most from AI content creation. There’s not much wiggle room in stories about financial performance or sports scores, so a robot can easily create a story that’s ready to publish.
The Associated Press announced it will use AI software to publish college sports articles about events that it wouldn’t otherwise be able to cover, giving more exposure to student athletes and smaller schools that don’t get full journalistic coverage.
Robots are also tackling breaking news stories and public safety announcements. With a few facts, AI writers can generate an article in minutes, alerting the public and letting the news organization publish the story first.
The good news for writers is that AI-generated stories aren’t necessarily meant for human consumption. When an AI-generated article sends a press release to national news organizations about a tsunami, the Washington Post and CNN send their reporters to get the full story. The only difference between a robot and human creation is that an alert can be sent out minutes after the tsunami strikes without risking human life to alert the media.
Outside the news world, businesses are starting to use AI generated content as well.
Citibank has employed AI services for more than a year to generate emotionally charged content for its email subscribers. Persado, the company behind Citi’s email generation, channels 19 different emotions and combines them with tens of millions of subjects lines. The results have been staggering:
Citi has seen a 70 percent increase in open rates and a 114 percent increase in click-through rates.
This proves that AI is just as capable of reporting facts as it is of persuading you to learn more about them.
Citi has countless email segments and customers in different parts of the funnel. While it has the resources to hire copywriters to constantly generate email copy for each of these segments, the work is boring and has a high turnover rate. Furthermore, the copywriters are mostly guessing about what appeals to customers.
AI technology doesn’t get tired of the work and actually improves with each email by learning what works and what doesn’t.
AI content creation is actually good news for businesses.
Most companies have been trying to personalize content for years through retargeting and auto-adding customer names to subject lines. However, the two most common emotions these personalization tactics elicit are anger and annoyance.
With AI, marketers are able to move past that. Algorithms learn when to push a sale and when to add an emotional hook. It’s more personal and less creepy.
Just like the trendy AI powered website builders that are popping up (i.e. PageCloud, Wix, The Grid), AI content creation isn’t going to replace the personal touch of creating effective content. Instead, AI writing conserves resources for small businesses that don’t have the time and patience to A/B test subject lines and website copy.
In the future, a small business can have a responsive website with regularly updated content that only takes five minutes to create. With all that free time, the owner can focus on developing new products or creating a better customer experience. Business will run smoother because time isn’t sucked up by testing.
If AI develops like a human and writes like a human, then can it talk like a human?
That’s what developers at Facebook and Slack have been trying to figure out.
In April 2016, Facebook released a chat bot API that’s meant to help companies provide customer service without a human touch.
The new chatbot, which will be channeled through its Messenger app, will provide basic knowledge like weather reports and sports scores but can also update customers with tracking numbers and offer customer service assistance. Similar to the Google RankBrain algorithm, Facebook’s chat bot is programmed to learn customer intent and respond using natural language.
In fact, if Facebook’s AI works out the way its developers plan, the Messenger app will replace customer service call centers entirely.
As Mark Zuckerberg cheekily put it during his April 2016 F8 summit,
“To order flowers on 1-800-Flowers, you never have to call 1-800-Flowers again.”
Slack, one of the fastest growing technology companies ever, is also building an AI chat bot to create a virtual assistant.
Right now, Slackbot offers tips and advice — along with jokes that Mark Zuckerberg would approve of — but developers are working to make it more personal.
CEO Stewart Butterfield wants Slackbot to schedule meetings, send follow-ups, and take over other mundane tasks that slow down the workday. However, he acknowledges that his company has a long trek ahead of it and claims that Siri is “nearly useless” even after Apple invested billions in the platform.
About 50 percent of iPhone users ignore Siri, which is an improvement from the 85 percent who ignored her in 2013.
More people are adopting hands-free virtual assistants in their cars, but Siri still seems like a fun toy instead of a necessary tool.
Most AI assistants still respond directly to questions instead of interpreting the needs of the user.
For example, Siri is likely to pull up a list of search results for any query, which defeats the purpose of hands-free voice search. Someone asking Siri a question on the highway still has to take his or her eyes off the road to search for a response.
For chat bots to improve and become essential personal assistants, they need to understand the intent of questions and even provide follow-up information.
If someone asks where the nearest barber is, Siri or Slackbot needs to find a highly rated barber and offer to schedule an appointment. Smarter virtual assistants would sync with the user’s calendar and make an appointment during an open block of time.
Robots aren’t full-scale solutions; they’re tools to make menial labor easier and more cost-effective. Small businesses need to embrace AI to better their customer service and overall experience.
If you view AI as a threat, then your business will get left behind. However, if you welcome these technological advancements and view them as opportunities to save yourself time and money, then your business will be on a growth trajectory.
To help you with this transition I have put together an actionable list of tips, tools and advice for making these robots an integrated part of your online marketing strategy.
Since Google’s RankBrain is thought to work similar to the Quality Score that is used in Google AdWords, it will be important for Google to see that your site is not only contextually relevant, but also extremely effective at retaining users. Metrics like click through rates on organic search results, site load time, bounce rate and average session duration become a lot more important than ever before. Focus on building a high quality website that wows your visitors.
Machine learning is an ever evolving technology that will get smarter each time. For this reason it makes sense to not just develop content around a specific keyword, but rather to think, implement and optimize around content concepts. For example, instead of one blog post on a certain topic consider developing a collection of posts that intertwine while covering all angles and aspects of the overall concept or theme. Create detailed, linked online resources that make you the authority for that concept or topic.
Unless you’re still sporting a flip phone, then you’ve already used AI as a personal assistant whether you know it or not. Siri, Google Now and Cortana are all popular AI assistants that continue to evolve and help us perform a multitude of actions. This same technology has quickly made its way into the business / marketing realm and can be one of your best resources if leveraged properly. For example, instead of wasting time and attention on scheduling a meeting you can use a service like X.ai to handle the back and forth for you. Or, automate redundant customer service situations with services like DigitalGenius. Embrace and utilize available AI tools so you can focus on improving your product or service and the overall experience that your customers receive.
Two of the most important positions in your company are sales and support. Your customers have their first and most interactions with these resources, which means that your business needs to be rock solid in both places. Instead of constantly churning through these resources or piling more onto your overhead it could make a lot of sense for you to leave some of these tasks to artificial intelligence. More specifically, services like Conversica and Automated Insights can help you find and qualify new prospects as well as automate some of the more repetitive tasks in your business like reporting or even content development. Consider using AI to more efficiently grow your team, improve your communications and provide better / faster experiences for your customers.
So, if you haven’t already, start thinking about how these different AI tools can be leveraged in your business. More specifically, think about how online marketing with artificial intelligence might be able to replace some of the more monotonous tasks that are eating up your valuable time.
At the end of the day, your focus needs to be on making the best possible products and / or providing memorable, positive customer experiences. Everything else is secondary and AI will allow you to do just that.
Need help strategizing the implementation of AI in your online marketing? Let’s talk!