Artificial Intelligence (AI) is being heralded as a game-changer – and rightly so. The combination of Big Data with improved processing power and machine-learning (algorithms that can learn from experience rather than explicit human programming) has the potential to affect every corner of our lives. From retail and healthcare to transportation and the internet-of-things, AI is poised to enhance and augment the human experience.
So, too, with marketing. But while some suggest that AI is changing the fundamental rules of marketing, I believe it is less the rules of marketing that are changing and more the tools and approach to understanding your customers.
AI is Not Human
While AI is good at logic, it isn’t necessarily good at “human.” And indeed, some have maintained that the focus of AI development should not be on what humans do well, but on the areas where humans typically stumble. AI is already better than you at finding patterns in (vast amounts of) data and making logical decisions based on those patterns; it’s unlikely, however, that AI will ever be better than you at orchestrating authentic, human relationships with your customers. Take humor, for example. It’s always been a staple of advertising but it’s the “last frontier” in AI because it’s so difficult for an algorithm-driven being to master the self-awareness, empathy, spontaneity and linguistic sophistication that humor requires. Nothing like a “that’s funny because…” logical explanation to kill a joke. Except for the instances where that’s funny. Because. Humor.
If anything, the past decade and metaphoric rise of social media has forced brands to become “more human.” Two-way communication based on authenticity and transparency effectively replaced the old-school focus on controlled brand messaging, because (as Marty Neumeier so succinctly put it), your brand isn’t what you say it is: it’s what your customers say it is. And boy howdy, your customers can talk – especially when they think you haven’t adequately met their needs. Brands have realized that to maintain a positive brand message that is at least partially in the hands of their customers, they need to authentically serve the needs of those customers.
Data is the New Currency of Authenticity
The first step to “authenticity” in marketing is knowing your audience; the second is to be relevant and focused on their needs; the third is to personalize their experience based on those needs. One-to-one personalization at scale has been marketing’s holy grail for years. Currently, however, using customer data to drive experiences across channels in real time is not a reality for most marketers. Only 7% of worldwide respondents to a Q1 2017 CMO Council and Redpoint Global survey said that they could “always” do this.
I’ve yet to meet a client whose problem was lack of customer data, however. From sales data to site analytics data to call center data to email profile data to browsing history data to social media data, there is a wealth of data points connected to your individual customers. Again, the problem isn’t lack of data – the challenge is actually the abundance of available customer data and how to effectively connect the dots to generate meaningful insights.
If you’re a marketer who’s faced this challenge, you know that it’s an overwhelming task for us mere humans. Normalizing cross-channel data to define a single view of your customers involves dealing with unstructured data, with different data structures, with multiple uncoordinated customers IDs, and with anonymous data not tied to any ID at all – usually across very siloed business units within your organization. What we find overwhelming, however, is what AI eats for breakfast.
CDPs Make Sense of Customer Data
Customer Data Platforms (CDPs) provide the structure to make sense of omni-channel customer data. Although they’ve been “a thing” since 2013, with the swift emergence of both machine learning and deep learning as viable marketing tools, the CDP segment is growing rapidly. A couple years ago there were maybe two or three CDP vendors; now there are over twenty.
So what exactly is a CDP? A CDP is a cloud-based data platform that aggregates customer-focused data from multiple, unrelated sources (both owned and external – so the data is also in multiple and unrelated formats with multiple and unrelated IDs), then cleans and unifies that data into a single customer view that you can make available to your other marketing systems. Unlike Data Management Platforms (DMPs), which are organized around anonymous cookies, CDPs tie data to identified individuals. And unlike CRMs, CDPs are designed to import large volumes of data from external systems that they then unify by matching different identifiers. Also worth noting is that CDPs are fully formed solutions designed to be owned and operated by marketing (rather than IT).
So what exactly does this have to do with AI? Two things, actually. The first is that many CDPs apply predictive analytics and machine learning (AI) to your customer data to recognize patterns and reduce data complexity — that thing that AI will always do better than humans. You can take advantage of this now. (Note, however, that machine learning and predictive modeling are not inherent elements of a CDP, which in its most essential form is simply a tool that provides a unified view of your customer from varied data sources. Different vendors take different approaches, so do your homework.)
But more importantly, a CDP can help you prepare for the application of deep learning (the “whoa” end of AI that organizes machine learning into increasingly complex application) to your customer experience strategies. At this point, you probably don’t yet know how exactly you’re going to leverage AI within your organization. What you do know is that the companies that figure this out will have a distinct competitive advantage. Organizing your customer data is a first step in this direction, because the fundamental rules of marketing haven’t changed – it all begins and ends with serving your customers. Having as much data as possible about them in a single database that is accessible to all the other systems in your marketing stack will put you at a distinct advantage to take advantage of AI.
Trying to plan for what has yet to be imagined can be overwhelming. I like how Carlos E. Perez over at Medium put it: “Deep Learning may look like alchemy today, but we eventually will learn to practice it like chemistry.” The game is changing, but you don’t need to be caught flat-footed.
