What is dynamic offer management? The traditional definition – the dynamic altering of the contents of an offer to better manage supply and demand – tells only half the story. Managing supply and demand or margins, such as an airline trying to fill seats, certainly justifies engaging in dynamic offer management, but that’s mostly for the brand’s purposes. A more modern definition also accounts for and elevates the customer as an equal. Through that lens, dynamic offer management is delivering contextually relevant information on the right channel at the right time, based on an individual customer’s behaviors and preferences.
Efficient management of inventory and margins, or delighting a customer with a personalized next-best action – an offer, a promotion, a message, etc. – all contribute to new revenue. The personalized next-best action, however, has the added advantage of optimizing customer lifetime value which, one could argue, is potentially far more beneficial than a one-time generic offer made to fill a flight, or to clear excess inventory.
Dynamic offer management is, at its core, is about maximizing relevance for an individual customer. And relevance, we know, is what today’s always-on, connected customer demands from the brands they engage with. Consider research from the Harris Poll sponsored by Redpoint, where 63 percent of consumers surveyed said that a personalized experience is now a standard expectation. Asked to cite examples of personalization they expect, 52 percent of consumers said it was when a brand sent special offers available only to them, 43 percent said it was when a brand recognizes them as the same customer across all channels, and 38 percent said it was receiving recommendations based on purchase or viewing history.
One-Size-Annoys All?
A discount that is sent to every customer not only fails to meet customers’ expectation for personalization, it has a high likelihood of introducing friction into a customer experience. We all know the drill; an inbox cluttered with generic “25 percent discount” offers for items you’re not interested in, or – worse – that you recently purchased at full price. In a traditional understanding of dynamic offer management, friction is a price brands are willing to pay for more efficient inventory management. Sure, the brand might annoy a certain percentage of customers, but it clears shelves to make room for its winter line.
The same approach in loyalty programs runs the risk of alienating a brand’s most loyal members. Sending a blanket 25 percent discount to everyone in the loyalty program may do more than introduce friction, it can also remove the incentive to remain a member. Why sign up if I’m not receiving special offers? Or why should I spend more to be a “platinum” member if I’m receiving the same offer as a “gold’ member?
A “dynamic” offer that is squarely focused on the brand’s bottom line, and not the end customer’s individual behaviors and preferences, does not align with rising customer expectations to be recognized as an individual.
Dynamic Offers and Real Time
To be successful, dynamic offer management must rest on a golden record, a real-time, continuously updated single view of the customer that includes data from every source and of every type. A golden customer record is built on advanced identity resolution capabilities that ensure an accurate customer match.
Knowing everything there is to know about a customer, in real time, is the secret behind dynamic offer management where the offer or action is personal and relevant to the customer’s journey at a precise moment in time, in whichever channel the customer chooses to engage.
Consider, for example, the difference between receiving an email offer for 25 percent off a winter coat you just bought or an offer for matching hat and gloves. The example highlights the need to have integrated customer data and real-time decisioning as key components of dynamic offer management. A brand may take an initial step of sending an email offer for the coat to a certain audience. But if a customer then goes online and purchases the coat – before opening the email – dynamic offer management requires the brand to have the ability to change the content of the email up to the moment the customer opens it. An offer for the matching hat and gloves might be determined to be the next-best action if the customer browses accessory items, for example. But if the customer takes a different action, dynamic offer management ensures that the next interaction will be in the precise context of the customer’s individual journey.
With a golden record and real-time decisioning, dynamic offer management does not even necessarily have to be an offer. Perhaps the brand’s intent is to have a customer download a mobile app. With open-time email, the brand can dynamically change the content to whatever the next-best action is – for that specific customer – based on the entirety of the customer’s behaviors. Every interaction is completely relevant at the precise moment the customer is interacting with the brand, for the channel they’re on.
Hands-Off Dynamic Offers with AML
A blending of traditional and modern dynamic offer management – pushing inventory or managing margins that also reflect an individual customer’s behaviors, preferences, and transactions – requires advanced analytics. Automated machine learning (AML) can handle endless permutations of offers and audiences. Code-free, self-training models that are optimized for a specific metric or business result can be programmed to strike the optimal balance between the customer and business goals, ensuring that the decision rendered accounts for any variable.
Automated machine learning eliminates the need for human judgment. Audience segmentation that relies on human input is inadequate to account for the vast differences in characteristics that make up a “look-alike” segment. Women ages 18-35 who live in the Midwest might all have a need for a winter coat based on a geographical segmentation, but that segmentation ignores other traits more indicative of a person’s characteristics, tastes and traits. A marketer may consult a spreadsheet to view a woman’s recent purchases, her size, and color and style preferences, but only AML can take this information in the form of a golden record that also includes viewing habits, channels and pages visited, time on page, social sentiment and other behaviors and actions to generate a real-time decision. Fleets of machine learning models are tuned a desired outcome, and when optimal conditions are met based on the various offers at hand, the customer’s behaviors and other variables, a winning model will render a decision optimized for maximum impact.
Dynamic offer management dressed as next-best actions are always in the cadence of the customer, delivered in the precise moment of a customer’s journey when they are most likely to positively influence and guide the journey to the desired outcome. A dynamic offer management approach that focuses on supply and demand may achieve short-term objectives. One that puts the customer first and consistently delivers personalized, relevant information is built for long-term success.
