Machine Learning will help the biggest brands boost ROI and consumer satisfaction.
Marketers have a data problem. Much of the explosion in data comes from consumer behavior and the digitization of human interactions. By 2020, we’ll be creating 45 zettabytes of data each year. For a sense of scale, one zettabyte is equivalent to 152 million years of UHD8K video! It’s difficult to wrap our heads around how much data we have and will have. This fire hose of information can offer huge benefits for marketers — if they can figure out how to manage it.
That’s where Artificial Intelligence, or AI, comes in. As data continues to expand, along with customer expectations of relevant real-time interactions, Machine Learning will become an essential tool. AI will help marketers connect the dots, as we develop algorithms that learn from real-time data in order to predict behavior and/or prescribe actions.
Trae Clevenger is EVP Analytics and Chief Strategy Officer at ANSIRA, the second largest independent digital and customer relationship management (CRM) agency in the U.S.
ANSIRA works directly with Fortune 500 companies such as Domino’s, Subaru, Panera Bread and Coca-Cola, giving Clevenger a firsthand view of the burgeoning AI landscape as it relates to digital marketing and advertising. Clevenger explains: “The issue is that most brands don’t know about the unique opportunities AI presents. Artificial Intelligence has so far mostly been applied on the operation side of businesses — very few companies are applying it in the marketing space.” Those who do, however, are seeing great results.
For example, one of ANSIRA’s restaurant clients interacts with customers across multiple devices and touch points (app, website from mobile or laptop, point of sale system, in-store kiosk, etc.). Their clients expect the company to know who they are in every channel, and to give them relevant deals and discounts based on their known preferences. When ANSIRA began managing the company’s loyalty program — using AI-driven offers, timing and interactions — the company saw a 2x-5x lift in transaction frequency and 2-3x boost in ticket price. That’s huge.
“Machine Learning is a powerful tool, but its successful implementation is about a lot more than computers and software. In order to truly apply AI at the enterprise level, organizations must combine nearly real-time data capture with real-time predictions. And then, crucially, those predictions must be activated across all the possible combinations of who, what, where, when, and how — using what we’re calling customer experience orchestration technology,” explains Clevenger. He concludes: “This means that companies need to hire experienced data scientists to help build and support the learning platforms necessary to create long-term business benefits.”
Clevenger warns that a lot of software vendors are promoting “built-in AI”, but what they really means is that the software has the capability to embed models and learning. You still need to have people who know how to feed data, generate models, and extrapolate information into actions. Clevenger notes that to run a one or two month “proof of concept test” can run $150,000 to $500,000. So, at least for now, it looks like only the biggest brands will be able to cash in on the AI potential of unlocking and using big data. When they do, the gains to both the brand and the consumers they serve look very promising.
This article was published on inc.com