Google CEO Sundar Pichai once said that Artificial Intelligence (AI) is one of the “most important things humanity is working on… more profound than… electricity or fire.” That quote has reverberated across every industry, and content marketing is no exception. But where exactly is AI going? What does the future hold? And perhaps most importantly for our purposes, how do you actually get started? Marketing AI Institute founder Paul Roetzer may have the answers.
Breaking AI Down
Roetzer likes to recall New Year’s Eve 2017, when he saw a series of tweets from a Bloomberg journalist listing the things that didn’t exist 10 years ago. We didn’t have Instagram, Snapchat, Alexa, or Slack. Plus, although the iPhone had only just been born, no one — not even Steve Jobs — could have fully predicted the impact that it would have. The same may very well be true for AI.
In simple terms, Roetzer says that AI augments what we’re capable of doing, at scales that we’re not. AI is already ubiquitous and something that we use dozens of times a day without even realizing it. From Facebook’s facial recognition software to tag people in photos, to Google Maps and Siri, it’s an integral part of modern day life. Not only that, with each passing day, AI gets more convenient and more personalized.
In AI, deep learning is trying to teach machines to think like the human mind. In content marketing applications of AI specifically, it’s about using Natural Language Processing and Natural Language Generation. So how is AI changing what we already do?
The 5 Ps for Getting Started with AI
Using AI to create better, faster, more accurate content is already a reality. The Associated Press reports that it has gone from publishing 300 earnings reports a quarter to 3,000 thanks to the help of AI systems. Meanwhile, in 2016, The Washington Post published high school sports stories, Olympics, and election coverage that once again machines had written. Clearly, the future is already here. If you have any kind of data-driven content, you need to get on board right now.
If you’re wondering how, here are Roetzer’s Five Ps to get started with AI:
First, you need to analyze your existing position. There are tons of specific software options that can give you useful information on where you are and where you want to be. You can capture information on your competitors, monitor the digital ad space, and determine how much others spend, how many ads they bought, and what channels they’re using. You can also map content to the buyer journey, identify keywords and topic clusters that can be measured against things you’ve already written, and predict how your content will perform.
Some software companies claims that machine-written subject lines for email out-perform human written ones 98 percent of the time. It may need 100,000 emails to learn from, but imagine a human trying to do that. To be able to create content at scale — across languages and across geographic markets, and to do it in the correct tone, using the right sentiment and adhering to brand guidelines — is a content marketer’s dream.
Using AI, content marketers can not only automate some data-driven content, but also surface data on what they should write about. But far from machines replacing human beings, Roetzer says this means using AI to enrich what content creators are already able to achieve.
These days a lot of marketing money is going into bots. That’s especially true of bots that initiate conversations with consumers at the right time. While we might rely on some study to tell us that the optimal time to send our emails is 2 pm on a Thursday, AI can do a much deeper dive. It can tell you exactly when a specific kind of customer will open a newsletter.
No matter whether you have 1,000 or 100,000 subscribers, humans will take a scattergun approach. They’ll set a rule that says send an email at a specific time and hope for the best. AI, on the other hand, can segment those customers at a whole different level. If, for example, just 1 percent of them open their emails at 11 pm on a Sunday night, then an AI system will make sure that the message arrives in those people’s inbox by 10.45 pm.
Should the play button be green or blue to get consumers to follow through? Some people will be more likely to click on one color, and some, the other. This is the kind of thing that humans can’t figure out with any accuracy. Nor should they be trying to. Especially when machines can do this more quickly, accurately, and efficiently on our behalf.
Using data, AI can write social media updates, figure out what image to use, what hashtags to include, and when to publish for maximum results. Current intelligent software using AI will recommend colors, surface hashtags, adjust the send date based on who you’re sending it to, and alter the channel to make each post as effective as it can be. Humans will never be able to do this so effectively, especially at scale.
We don’t just want to create perfect content, we want to be able to predict what will be the most successful content. Just as companies like Amazon and Google make data-based recommendations of what we’re searching for, we need to be able to create content that will most appeal to consumers based on what we know about them. Collating and interpreting the relevant data is time better spent by machines, rather than writers who craft the essential message.
Moving Forward with AI Today
To get on board with today’s AI, and move with it into the future, Roetzer suggests that content leaders look at the things that they and their writers do every day. Look at things that are manual and repetitive and require data to either make decisions or create and produce content, he says. Once identified, content managers should consider the AI capabilities within their existing marketing technologies and explore the potential of emerging solutions. The bottom line, he says, is that when you discover ways to use software more intelligently, you’ve got more time to create intelligent content.
Want to hear more from Paul Roetzer? See his video below from our recent Content Connections 2018 user conference.
Learn How Top Businesses Do Content Development
Acrolinx recently conducted a survey of 250+ content professionals from around the world to better understand their content creation practices, goals, and challenges. Download the report below to get recommendations for what every content professional should be doing in their content operations.