Next best action (NBA), also known as next-gen engagement, refers to marketing models that use predictive analytics and machine learning to recommend, in real time, actions that are likely to be taken by a customer based on the customer’s profile, previous actions and needs. Among all possible potential actions, NBA is that which most benefits the consumer, delivering to them what they need and desire, and providing to the brand access to the best customers in real time.
It’s personalized marketing in hyperdrive, in other words.
How health marketing benefits from NBA
The issue of privacy today figures largely into mapping consumer journeys, particularly in health marketing. In healthcare, that journey encompasses questions the patient asks, such as, What are my symptoms? What do they mean? What’s my diagnosis? What are my treatment options? How will they impact me, my family and my finances?
Consider, by way of comparison, the consumer journey in the automobile space. At first, the customer is not completely convinced of what kind of car she wants, so she searches the internet to learn more, then eventually progresses to a visit to the dealer for a test drive. She’s now in the space to purchase. It is a cleaner, more linear journey.
So there are obviously much more sensitive elements of the healthcare journey than the car-buying process. Clearly, marketers are not supposed to be privy to a conversation a patient has with his doctor. But the marketer can surmise to an extent, and from a higher level, that the patient has tried different treatments and is in the space to learn about other drugs.
When we talk about enriching the AI, you can go very deep. You're not just feeding the machine with your basic demographics data, but also actually introducing where a person sits in his patient journey.
Refining a model of different inputs
But the truth is, marketers—whether trying to reach consumers or healthcare patients—don’t fully understand that NBA is not just about machine learning and artificial intelligence, or even data. There is, rather, a huge piece of it that is just about getting access to the right data to feed into the technology. You can have the most sophisticated algorithm around, but without the right data there’s nothing to which to apply all that fancy machine learning.
The question is, Do you have the right data and technology for all the different touch points along the consumer journey?
If we look at long-term marketing goals, we must look at the different elements we want to personalize and figure out how to personalize them correctly. What is the optimal creative to reach the target? What are the best touch points? What is the optimal medium—is it YouTube or an email? And is this the right moment to try to reach them?
If you’re a brand, you know your product, you know your clientele, you know which marketing tactics have worked in the past, you know what drives consumer behavior and which audiences you aim to reach and what drives your own success.
All that amounts to a hypothesis. What NBA does is constantly refine a model of different inputs around that hypothesis to ensure the most optimal marketing plan is constantly being presented.
The need for data integration
How does this differ from traditional marketing? The old-school way would be to look at a range of different audience segments and say, I know this message works with this segment and I understand what the customer journey there is. So the marketer maps it out hypothetically and starts communicating with the customer based on that. Then AI takes all that data in and figures out what’s working best. Along the way it gets smarter.
In contrast, next best action is where the brand says, I’ve been running campaigns and I can see the results, what’s working and what isn’t—but I now have developed some real understanding of why certain tactics are working and others aren’t. I use that knowledge to not only build what I see as the customer journey, but also specific details. Where the machine takes over is in finding the cohorts and figuring out how the puzzle pieces fit together to maximize the efficacy of a marketing plan. The best part is, it takes a matter of minutes or days as opposed to the many months it would take a person to piece it all together.
With next best action, the machine is constantly trying out the next hypothesis, constantly evolving which steps it thinks are best and constantly learning from those decisions.
The challenge for marketers remains: That the data required for effective personalization is trapped in silos. The solution? Integrating customer-data and data-management platforms, augmented with identity-resolution platforms, to unify data and make it available across channels for activation. Data must be centralized and made available so activity in one channel can immediately support engagement in another, in real time or near-real time.
Here’s another challenge: Decisioning logic resides in individual, channel-based, black-box systems—or it does not exist at all. This results in a disjointed experience, whether for customers or healthcare patients. The answer is creating an integrated decisioning engine that uses machine learning and AI models to score various propensities for each person.
At PulsePoint, we are continuously refining our underlying model inputs to make sure the healthcare options presented are dynamic and better prioritized on an individual level.
At its base, of course, is data—a whole lot of data. Getting a more nuanced, personalized view requires mining and synchronizing more extensive and unique types of data. By using data personalized to the patient, we can help determine not just a next best action, but their next best action, and do this proactively.