June 5, 2018
Marketing AI: all hype?
Have you wondered if you’d benefit from one of the new marketing tools powered by artificial intelligence? Some of these platforms make claims that sound almost magical, and it can be tricky to sort out whether they’re worth the investment. Here’s a look at three of the major players in marketing AI, together with an overview of their unique capabilities — as well as a glance at which claims are over the top.
Einstein is the AI platform created by Salesforce and launched in Fall 2016. It has grown in sophistication since then, using customer data to automatically generate models. These models are continually improved by the AI, which analyzes the history of data and decides which factors are most accurate at predicting the behavior of individual customers. As the unit receives more information, it learns which of its models need adjusting, without the need for any intervention by developers.
Salesforce recently announced that its Einstein engine is now delivering over 1 billion pieces of information to its customers each day. It ranks lists of sales leads and puts information at one side of the screen to show each prospect’s probability of deciding to buy. For marketers, Einstein sorts email lists and says which recipients are most likely to open a given message. Salesforce’s director of product marketing Ally Witherspoon gives an example in Wired of a solar energy supplier discovering through AI that a person’s chances of buying are influenced by the pitch of their roof. In the future, she envisions, satellite photography could tag prospects’ homes according to the geometry of their roof.
Einstein is also in a partnership with IBM’s Watson, integrating its own CRM data with Watson’s insights. This wider reach offers access to customer data sources beyond what Salesforce itself is able to collect.
Motiva AI Cloud
Motiva works with Oracle’s Eloqua to optimize messaging through its ability to characterize audiences. In the words of Chris Diehl, Motiva’s CTO, the purpose of the AI is “removing the need to manually define the relevant population.” Diehl points out that by using artificial intelligence, marketers can discover new affinities and associations and “uncover meaningful populations that exhibit shared content preferences.” In other words, the AI gives new insight into whom to target and which target populations can be grouped together.
The real-world example that Diehl describes is very similar to that offered by Witherspoon when she talked about Einstein. Diehl mentions learning the external web browsing behavior of a marketing audience, then using that data to sort and rank this audience. Once the Motiva AI platform knows these people’s messaging channel preferences, it will advise the marketer about how to customize their messaging. In one example, a healthcare provider used Motiva to vary and improve the messaging sent to patients, and achieved a doubling of click-through rates.
Adobe Sensei is an interesting addition to the AI mix, because unlike Salesforce or Oracle Eloqua, Adobe didn’t originate with CRM and marketing. Instead, the company drew on its deep graphic expertise and entered the AI universe through its interest in recognizing and manipulating visual data. This exploration eventually led Adobe to capitalize on the fact that logical patterns lend themselves to machine learning as readily as do visual patterns.
Fortune magazine says, “Sensei pores over tons of data (the more the better) to detect patterns and present results in a visual way.” While Adobe offers much the same customer experience options as the other two marketing AI systems mentioned here, it plays up two specific virtues: information security and cross-device access. Its “Cross-Device Co-op” program lets brands who use other Adobe marketing products recognize their consumers regardless of what type of device those consumers use. Adobe points out that marketing spend will decrease, because companies can “focus on people, not on devices.”
A number of banks currently use Sensei. If anonymous users visit the bank’s website, the AI platform can make suggestions about products to show them, even if the user doesn’t have a profile with the bank. HSBC used Sensei to highlight a specific product and saw a 109 percent increase in customers reaching that product.
But Will AI Really Help You Market Your Products?
Automatic customer model generation is flashy, but the amount of time it will save you depends on your individual needs. A platform like Salesforce allows businesses that aren’t Google or Facebook to simply purchase AI power without having to customize it and teach it what’s important to that particular business. That efficiency can definitely come in handy if you need help identifying your audience.
Eloqua Motiva can develop and dynamically revise models for messaging prospects automatically, so that the human marketer doesn’t have to waste time testing out which message will work best with which group.
On the other hand, when Einstein itemizes the factors it uses to score prospects, that transparency can either be helpful or distracting. And it’s definitely hype when Salesforce puts out a press release saying that by 2021, “AI-powered CRM activities ‘could’ increase global business revenues by $1.1 trillion and create 800,000 net-new jobs.” Salesforce customers alone, it points out, will account for $293 billion of those jobs. While it does seem promising that AI will increase the number of jobs overall, due to streamlining the time-consuming processes of segmenting and testing audiences, sky-high predictions are not enlightening.
Sorting Out What AI Can Offer You
Tech developers are understandably enthused over the capacity of AI, and marketing is indeed a practice that needs to be — at least partly — data driven. Fifty-one percent of marketers currently use some form of artificial intelligence, and no human being can handle data with the finesse of AI. But the claims made by some of these companies go a bit too far; they imply that you can just toss an AI platform into your marketing campaign and passively let it work its behind-the-scenes wizardry. As Motiva’s website seductively promises, “Turn it on, get results.” No AI platform can synthesize the intuitive insight of a skilled marketer.
There’s an elusive sweet spot between the creativity of a human marketing professional and the efficiencies offered by artificial intelligence. AI can be a useful tool, as long as you identify your goals ahead of time and have a clear idea of how you’ll use intelligent data manipulation to streamline your daily tasks. Certainly the platforms we’ve mentioned above can save time by making sure you’re speaking to the right audience, delivering the message they want to hear via the channels they like best. Furthermore, each system focuses on differentiating themselves from their competitors by delivering something unique. In our upcoming posts, we’ll take a deeper dive into what each of these three brands have to offer. We’ll examine their specific benefits and help you determine which one would be most applicable to your particular situation. We’ll also sort through the three brands’ own verbiage and clarify which of their claims are actually meaningful and which ones are just … well … marketing.