This article was published on 10/23/2017 in VentureBeat.
Since the invention of mass media, arguably, the primary focus of marketing has been to increase its level of personalization. Marketers constantly seek more targeted audiences, and strive to deliver messages that speak more directly to their diverse audiences. So, it's no surprise that AI and machine learning — with their ability to predict consumer behavior and make personalized recommendations on-the-fly — has captured the attention of the marketing world. But the elephant in the room is that advances in machine learning have far outpaced most marketers' ability to harness them
Unfortunately, this inability to personalize the customer experience is a huge missed opportunity. Customers now expect a tailored experience, including customized recommendations and a personal touch. And customers are willing to reward companies who provide it. According to Gartner, "By 2018, organizations that have fully invested in all types of personalization will outsell companies that have not by 20%." Even more alarming, customers are increasingly likely to dump brands that don't offer personalization. According to a 2016 Salesforce study: "… more than half (52%) of consumers are likely to switch brands if a company doesn't make an effort to personalize communications to them, 65% of business buyers say the same about vendor relationships."
But, there is a silver lining. Right now, AI and personalization is not just a challenge for some companies, it's a challenge for most companies. According to Salesforce, only about a quarter of all business leaders have confidence in their organization's ability to define an AI strategy: "AI interest persists, but many are grappling with what it means for their business. And while AI is on the tip of marketers' tongues, roadblocks still exist." This means that the opportunity has never been better for forward-thinking marketers to use AI to drive a competitive advantage for their organizations.
The good news is that help is already here. Cloud computing providers like Microsoft, Google and Amazon are making huge investments in machine learning, including teaching mere mortals like us how to use it. These providers are banking on the belief that AI will drive the consumption of cloud services. Similarly, a large number of marketing technology vendors are starting to bake machine learning into their products. So, although personalization is not quite turnkey yet, the opportunity has never been better for marketers to start taking advantage of AI.
But machine learning and personalization are such big topics, where do you begin? Fortunately, you don't need to be an expert to start taking advantage of AI in your marketing. The key is to start doing something, anything to build up your marketing org's working knowledge of AI. Start small, keep it simple, measure your results, and maximize your return on learning. In this article, I've outlined 4 ways to get started.
But first, some fine print. I’m using the concepts of AI, Machine Learning and Personalization somewhat interchangeably here, although in reality, they are different things. Personalization is an outcome. It is a tailored experience based on data about the user. In contrast, AI and machine learning are methods. They represent one way—and certainly not the only way—to achieve personalization. The goal for marketers should be to improve marketing performance by providing more relevant experiences. Whether this is achieved through AI or not is a secondary consideration.
1. Use A/B testing tools for permanent personalization
Traditionally, A/B testing tools like Optimizely and Adobe Target were used to test variations of page design. But more recently, these vendors have beefed up their offerings to provide always-on personalization for both websites and mobile apps. This route to personalization offers three big advantages for marketers. First, it's very easy to connect these tools to customer segments defined in analytics platforms like Google Analytics or Adobe Analytics. And by using these behavior-based segments, you bypass the need for large, IT-driven, data integration projects altogether.
The second advantage is that most A/B testing tools have rich WISIWYG editors that give marketers the ability to update their sites or apps directly. This means that marketers can publish different messages to different segments themselves, without IT support. Third, if you make multiple variations of creative for each segment, A/B testing tools can use machine learning algorithms under-the-hood to find the optimal variation for each segment while the campaign is in-flight.
2. Talk to your customers directly with a chatbot or voice-activated app
The ultimate personal experience is simply a conversation between your customer and a knowledgeable, friendly person within your organization. Historically, this type of interaction has been fulfilled by store clerks, sales reps or customer support. However, with the rise of AI technology, it's now possible to simulate a one-on-one dialog with digital conversational agents via chatbots or voice-activate applications. For example, using existing chatbot platforms like Wit.ai (Facebook), LUIS (Microsoft), or Chatfuel, you can build a chatbot on your website or mobile app that answers your customers questions, and gets them the information they're looking for quickly and efficiently. Even more interesting, these chatbot platforms can integrate directly with existing messaging platforms like Slack, Facebook Messenger, Kik, and WeChat (and even plain old text messaging) which means you can interact with your customers directly via the apps they already have on their phone.
Voice activated apps like Amazon Echo/Alexa skills extend the concept of digital dialog even further, by allowing a natural voice conversation to take place between your customer and your brand. By building an Alexa skill (a "skill" is an Amazon term for voice apps), you are literally one voice command away from your customer in their home, car and even at work.
When you’re starting your first chatbot project, just keep in mind that the underlying natural language processing algorithms that power chatbot platforms are still evolving. Set reasonable goals for your first iteration. The more things you want your chatbot to respond to (these are called "intents"), the more difficult it will be to create a frustration-free experience for your customers. Keep your goals simple at first and the number of intents small. Then, once you have your chatbot dialed in, start experimenting and expand the scope of your chatbot. And don't skimp out on training it, because somebody has to feed the chatbots.
3. Select AI-enabled MarTech tools
Perhaps the easiest way to start taking advantage of AI and personalization is simply to select marketing technology vendors who are serious about machine learning, and are building AI into their platforms. While this may not always be the cheapest option (the price tag for a full-featured platform rollout can get expensive), it is certainly the path of least resistance. Three of the leaders in MarTech AI right now are Salesforce, HubSpot and Adobe. Salesforce has made a dizzying array of AI-related acquisition over the last several years, culminating in the rollout of their branded-AI technology, Einstein, last year. Einstein includes a host of AI-powered functionality, including predictive lead scoring (identifying which leads are most likely to convert), personalization and product recommendation.
Upstart CRM platform HubSpot has also been investing heavily in AI, with a strong focus on content and digital analytics. Their recent acquisition of chatbot platform Motion.ai is a great example of how HubSpot continues to adapt with lightning speed to emerging marketing trends.
But the 800-pound gorilla of MarTech remains Adobe. Personalization and the rich use of customer data has been at the core of Adobe’s marketing cloud for years. More recently, Adobe enhanced their core platform with their own branded AI technology, Sensei. Sensei now fuels a range of machine learning functionality under-the-hood. For example, Sensei provides anomaly detection (identifying when statistically abnormal traffic spikes occur and explaining why), content optimization (showing different content to different customer segments based on which is the most likely to perform), and optimizing media spend (recommending how much to spend across each marketing channel and predicting future return). The breadth of the Adobe platform, plus the ease of data integration between products, combined with the growing AI capability of Sensei, makes the Adobe marketing cloud a MarTech powerhouse.
4. DIY machine learning through APIs
But you don’t need to rollout an enterprise marketing platform to start doing personalization. More often than not, marketers don’t have the budgets to support a large platform rollout, or they are stuck with a deeply entrenched legacy platform that doesn’t provide modern personalization capabilities. For marketers who fall into this bucket, another approach is to simply start building your own DIY personalized digital experiences using off-the-shelf machine learning APIs. APIs (or Application Programming Interfaces) are toolkits for developers that provide deep functionality through easy-to-use interfaces. APIs work with multiple programming languages on multiple platforms, making it a great option for people working with legacy digital technology.
Cloud providers like Google, Microsoft and Amazon have abstracted the complex, PhD-level mathematics that underlie most machine learning algorithms into intelligent APIs that your typical web or mobile developer can use. This means that you can leverage your existing digital team to start building AI-enabled personalization into your digital experience today, without hiring data scientists.
Microsoft, one of the leaders in this space, has a suite of over 30 machine learning APIs within their Cognitive Services offering. These APIs handle everything from computer vision, to natural language processing to production recommendations. Their Custom Decision Service API, for example, uses reinforcement learning to study your customers preferences and provide personalized content on-the-fly.
Similarly, Amazon’s cloud platform, AWS, provides machine learning APIs for image recognition, language recognition and generating life-like speech, enabling you to literally speak to your customers digitally. Although Amazon’s suite of AI-powered APIs is not as broad as Microsoft’s, they have the advantage of Amazon’s deep, real-world, working knowledge of AI gleaned from their behemoth e-commerce platform. So, what Amazon lacks in breadth, they make up for in depth and scale.
There are many different options available to marketers who want to start incorporating AI and personalization into their digital experiences. And while the rapid rise of AI in marketing can feel a bit daunting at times, it’s critical to start dipping your toe in the water now, so you can begin building up your organization’s experience with AI. A small investment today will yield a large competitive advantage tomorrow, as your organization moves toward becoming one of the AI haves (not one of the have-nots), and your customers start gravitating to your digital ecosystem because, well, you just seem to "get them" more than the competition.