Nearly every industry today is abuzz with the promises of artificial intelligence. AI’s applications span multiple types of business, touting benefits of automation, machine learning, and – more broadly speaking – intrinsically transforming the way the world of enterprise works.
However, there’s a little niggle in the nomenclature that plagues AI when we apply it to marketing: artificial. If AI is the marketing revolution we’ve all been waiting for, how can we transform it from something artificial into real, actionable ways of improving the way our company’s marketing works?
Fear not, there’s hope. Amidst the hype of artificial intelligence in marketing is a solid foundation of solutions that not only preach the benefits of AI, but actually deliver. Let’s line up the most promising of these solutions, and determine which of them could be a real-life game changer for the business world.
Artificial Intelligence and Written Content
This is an area of AI that will immediately raise red flags for some. How can a computer create a thoughtful, informed, innovative piece of written content that rivals that of a human being?
Simply put, it can’t. AI is years, probably centuries, away from being able to develop engaging, thought-leadership content for a human audience.
Yet all is not lost: if it’s fact-based content you’re after, this is where AI has been earning its stripes lately – and in a big way. Companies like Narrative Science (makers of Quill) and Automated Insights have developed technologies that take a set of data surrounding a particular subject and, using a proprietary algorithm built with a specific set of vocabulary, produce natural-sounding written content. Both Narrative Science and Automated Insights have been around for a few years, but they’ve spent their time refining their systems to their current states. And we have to say, it’s pretty darn good.
For example, Automated Insights’ AI technology was used to produce this story about a Major League Baseball game. Here’s a snippet:
Cristian Alvarado tossed a one-hit shutout and Yermin Mercedes homered and had two hits, driving in two, as the Delmarva Shorebirds topped the Greensboro Grasshoppers 6-0 in the second game of a doubleheader on Wednesday.
In a fraction of the time it would have taken a human to write this content, Automated Insights’ artificial intelligence wrote an article that was of the same quality, if not better, than a human sports reporter.
Marketing AI like that from Automated Insights and Narrative Sciences poses huge potential benefits for companies. That’s not to say that machines are necessarily better at producing written content, but rather that human workers’ resources should be utilized in areas where technology is currently – and perhaps will forever be – lacking: creation of ideas, opinions, innovation, new ways of thinking, the Achilles’ heel of computers. For starters, humans are more expensive and slower than machines. Lest we forget the lessons of George Orwell’s 1984, there are parts of content creation that can be left to machines. When written content is merely a regurgitation of a set of facts, marketing AI can save both time and money for companies.
Another way AI in content offers huge potential benefits is in SEO. A pillar of SEO strategy is to post relevant, fresh, keyword-rich content on a regular basis. AI marketing technology uses data from recent news stories and weaves it into an article while employing a weighted set of keywords focused on SEO strategy, which can then be posted directly to a website. AI therefore gives businesses a constant stream of website content that the bots at Google are sure to love.
Machine Learning and the Sales Funnel
Discussing the ins, outs, ups downs, and everything-in-betweens of the sales funnel is the bane of many marketing and sales teams’ existence. But what if marketing AI could help these teams navigate their sales funnel – say, for example, predict both the likelihood of a given lead converting to a sale and give an estimate of how much the sale would be worth, based solely on the lead’s behavior? Or if you could predict that one of your current customers was about to start spending more or less before even they know that? For the right type of business, machine learning with predictive analytics could be the most significant game changer to date.
Big data has been available for quite some time now, but we’ve been waiting for the software tools that will allow us to utilize this data. And that’s where the predictive analytics come in. The data set provided for machine learning is the key determinant to how well it can work: machine learning is used to create propensity models that assess a given lead’s behavior and, depending on the specificity of the data set on which it’s built, determine how likely it is for that lead to become a sale. From there, human resources from the sales team can either be allocated to the lead in order to nurture it through to a sale, or the lead can be abandoned so that resources don’t have to spend time chasing a prospect who is unlikely to convert.
One big player in predictive analytics is IBM, which has two main tools – SPSS Statistics and SPSS Modeler – designed to have a large breadth of applications in multiple parts of a large enterprise. SPSS Statistics is optimized for managing large data sets and producing advanced analytics models for a company’s broader marketing plan, where SPSS is action-focused, building models that help businesses make important decisions or realign their primary foci. Another company that offers predictive marketing analytics is Optimove, whose approach is somewhat more customer friendly and easy to understand, and better suited to SMEs.
This is another area where predictive analytics can pay off in dividends for a marketing team. Propensity models are created to track a given customer’s behavior in relation to their likelihood to convert, as we discussed above, but dynamic pricing introduces an additional opportunity to convert. Dynamic pricing offers a product at a discounted rate when the propensity model predicts it is necessary to get the sale. By doing this, only some customers are offered a product at a lower cost, therefore increasing the overall profit for the company by not offering the discount to everyone, as well as maintaining a high rate of conversion for customers who would otherwise have moved on to a competitor. It’s a bit like those website pop-ups that are triggered when a user begins to move the mouse toward the back or close button, presenting a desired CTA like a newsletter sign-up or discount. Dynamic pricing goes a step further and bases the incentive off a more specific set of behaviors, increasing the overall probability of a sale.
IBM is also a big player in this space, however again is focused at enterprise-level businesses. For companies with both brick-and-mortar stores and an ecommerce setup, dynamic pricing can coordinate prices among all touch points to influence online prices and physical stores. For companies that are publicly listed, it can also take into account market fluctuations when adjusting pricing. Omnia is a company that takes a connected network of sales locations, both online and offline, which they call “omni-channel profit”.
Chatbots are another product of marketing and machine learning designed to improve the efficiency with which your customer service team operates. However, they’re not as expensive or difficult to create as one might assume, and as such are available for many businesses to take advantage of, even without a massive budget.
Facebook has developed an easy-to-use development feature to help businesses create their own chatbots on the Facebook Messenger platform. The chatbots created are capable of designing interactive and engaging CTAs, then sending them to customers along with text and images, as well as staying on-brand with a custom welcome screen and invitation to start a conversation.
Even if your business just uses a chatbot as a gatekeeper in order to channel customer inquiries to the appropriate team member, you’re offering your customers a higher and more efficient quality of service while allowing specific allocation of valuable human resources to a place where they will provide the most value for your business.
This is one area of AI that you’ve probably already noticed, although its application to digital marketing is somewhat out of the box. Possibly one of the most impressive abilities of search engines like Google and Bing is their ability to use AI to know determine you are attempting to search for, even when your search term isn’t an exact match for your intended results. When it comes to marketing, this is a game changer for SEO: soon it will not be enough to optimize pages with keywords alone, but the focus will have to shift to writing about topics more broadly, including both keywords in multiple related formats (“AI in marketing,” “market with AI,” how to use AI in your marketing”) as well as related phrases and ideas (“machine learning and marketing,” “AI and the sales funnel”, “how to automate sales conversions”). Weaving these tactics into your content will give you a better shot at making page one as artificial intelligence solidifies its place in search engines.
* * *
At the core of all these types of artificial intelligence is an easier, more intelligent way to do work. Marketing will never be an entirely computer-driven task: humans will always be required to read the subtle nuances of the market and adapt accordingly. However, many of the processes that are part of a marketing department’s daily routine can be automated using fewer human resources than before, freeing up those resources to perform human-only tasks. We’re starting to discover that AI is not as artificial as we previously thought, and that real results can be delivered today.