Data•Apr 29, 2020
Apply AI & ML to Your Predictive Next Best Actions and Meet Your Customers Where They Are
If you are reading this, I would wager that two things are true of you:
You appreciate when someone has used everything they know about you to inform what would make a great gift.
Your time and attention are some of your most scarce resources.
These human traits are innate to all of us and render it crucial to acknowledge the importance of preferences and behaviors in delivering the best possible customer experience. In recent decades, as the variety of marketing channels and volume of content have grown, consumers’ attention spans have shrunk. In today’s omni-channeled, data-driven, and instantly-gratified world, organizations are competing not only on product and price, but also on memorable customer experiences. Brands can retain their audiences’ attention, earn their loyalty, and help them feel cared for only by effectively leveraging customer data and focusing on meeting customer needs. Once the right data is available and has been consolidated into a single customer profile, companies can apply analytics to respond to customer behaviors with the “Next Best Action.”
What is the “Next Best Action?”
“Next Best Action” (NBA) typically refers to predictive and prescriptive machine-learning algorithms that help organizations find patterns in the way customers respond to different touchpoints and then determine which actions are most likely to lead to conversion. As customers’ behavior changes, their behavioral data feeds the algorithm, which recalculates the probability of conversion for different touchpoint options. NBA decisioning chooses the best channel to deliver the best message at the best time.
What are the benefits?
Using NBA decisioning as a means for 1:1 marketing can help your organization serve your audience in a relevant and contextual manner by evaluating their interests and behavior. In this way, you can avoid dispatching touchpoints excessively or prematurely and, consequently, coming off as aggressive or out of touch. In addition to focusing inbound marketing on drawing in new leads, organizations can implement NBA algorithms that nurture existing relationships, improve customer perception, and drive behavioral shifts by providing personalized experiences. This type of frequency management also helps optimize marketing spend by limiting wasted touchpoints with customers.
What are the prerequisites for a successful implementation?
The barriers to overcome before leveraging NBA are the same barriers that impede any organization from achieving the full potential of their customer data. For example, as Credera’s Greg Gough explains in this blog post about the importance of robust customer analytics, customer insights are often spread across product- or channel- based silos. To realize the full potential of customer data, a company’s organizational structure must identify who has ownership of customer lifecycle metrics.
NBA decisioning requires incorporating data across the entire customer journey into a single 360-degree customer view. Whether this journey data tells you about a customer’s recent web visits, the emails they’ve received, or their latest interaction with a call center agent, the data comes from a wide variety of sources, which has implications for customer identity management. Customer data platforms (CDPs) are toolsets that can ingest and “stitch” data that would not integrate as easily into a traditional data warehouse and could result in a single customer having multiple “identities.” Once the organization has achieved this single customer view, CDPs can calculate scores for different customer metrics, such as customer lifetime value (CLV), based on business rules. NBA algorithms can be applied across channels to turn these calculated customer attributes into 1:1 marketing.
Of course, rich customer data won’t get you very far without the right content. After all, delivering the NBA means using the best channel to deliver the best message. Therefore, an organization must have personalization-ready, reusable content that is flagged with metadata taxonomy.
What are some interesting use cases of NBA Decisioning?
Marketing cloud suites, such as Adobe and Salesforce, have built-in artificial intelligence (AI) and machine-learning technologies with applications across various lines of business, from sales to service to marketing.
For example, Adobe Sensei‘s Auto Target (“One Click Personalization”) feature uses machine-learning models that learn what kind of content layouts resonate with consumers, which empowers the marketer to match each customer with the experience most likely to please them. Furthermore, algorithmically guided tools can help sellers find the right cadence for contacting prospects and show sellers what kind of messaging has the highest likelihood of closing deals.
In another example, Salesforce Einstein uses Next Best Action technology to help sellers reach decisions on what to recommend, when to do so, and what channels to use via an embedded Einstein Prediction Builder on customer contact pages. NBA could also benefit contact center agents, who often do not have visibility into the channels customers use before they reach out for help and, therefore, risk frustrating customers by asking for data they have already provided. In this example, AI-guided interactions would prescribe the appropriate path forward based on knowledge of what has solved similar issues in the past.
You may be wondering at this point whether you really need to be running machine-learning to inform your customer interactions. Sometimes, the logic for determining and pushing an NBA, like a paid social ad, is as simple as tagging a customer with an attribute if they’ve satisfied a set of conditions, which doesn’t require machine-learning. However, in cases where the marketer cannot wait for a set of conditions to be met and needs to know at any given moment which social media platforms work best for a customer, they could be running a machine-learning algorithm to make this determination. Take this example of the benefits of AI-enabled marketing over rules-based personalization: in an email campaign supported by Symphony RetailAI, a provider of AI-enabled decision platforms, a retailer tested Symphony RetailAI’s marketing solution against its existing rules-based personalized offers. After just three weeks, the retailer realized a 34% lift in redemption from the customers targeted by the AI marketing solution.
Try It Yourself
Leveraging predictive and prescriptive analytics that treat each of your customers as unique individuals means advancing them along various stages of their engagement journey. Next Best Action algorithms can help get your customers across the conversion line with the fewest interactions and lowest marketing spend. Keep the following prerequisites in mind when considering whether your organization is ready to adopt:
Get your marketing teams out of channel-based silos
Stitch your data into a 360-degree customer view
Have personalized, tagged content at the ready
These changes won’t happen overnight; adopting NBA tools requires a defined roadmap to shift from a product-centric to a customer-centric marketing strategy. However, despite the complexity of predictive technology and its reliance on pre-existing data collection and management, 34.5% of sales organizations expect algorithmic-guided selling to become more important to their customer relationships within the next two years, according to Gartner. It is clear that AI-based recommendations are the future of customer engagement.
Ready to empower your 1:1 marketing efforts with NBA decisioning? Reach out to our MarTech experts at firstname.lastname@example.org.
- Consumer Behavior
- Machine Learning
- Machine Learning Applications
- Artificial Intelligence
- Predictive Analytics