Break down how your customers make decisions with strategies that enable the elusive goal: N=1 marketing
ZS is offering a new approach to achieve every marketer’s dream — nano-segmentation to support omnichannel marketing at an N=1 physician level.
How is nano-segmentation different than segmentation or micro-segmentation?
· Segmentation identifies a small number of segments for strategy development (e.g., positioning, messaging, etc.) and subsequent tracking.
· Micro-segmentation combines different segments (e.g., specialty, market decile and channel affinity) to create target lists.
· Nano-segmentation leverages large amounts of individual customer data to deliver the right message at the right time for omnichannel planning and orchestration.
Most orchestration engines use machine learning (ML) algorithms to conduct “look-alike” analyses that implicitly define nano-segments. We propose a segmentation scheme that will both improve the “look-alike” analysis and increase transparency.
Transparency remains critical throughout this process. Current approaches are designed to maximize engagement through opens and clicks, but these strategies offer little information on what messages are missing. They also do not provide a roadmap of how messaging should evolve over time.
In contrast, marketers must continuously update nano-segmentation as new data becomes available. The right message and the right time to reach a certain customer is always shifting.
Why isn’t micro-segmentation enough?
The main drawbacks to micro-targeting are:
· This method divides segments based on secondary behaviors like prescribing trends or channel preferences, but it rarely provides insights into the motivating factors behind those behaviors.
· Analytics typically captures behaviors at a high level — for instance, a customer may be more likely to open emails with efficacy messages — and lacks the nuance needed to customize messaging.
· Organizations do not update behaviors frequently, so the recommendations remain too static for dynamic campaigns.
As a result, micro-segmentation works for developing custom campaigns, but it is too effort-intensive to scale beyond a few campaigns.
How will nano-segmentation get us closer to N=1 marketing?
We need to expand the information we have on physicians and how they make decisions. The traditional mix of Rx behaviors — like deciles, growers, maintainers or decliners — combined with market research personas and channel affinities is not sufficient. We must expand the amount of customer data we collect to feed our predictive algorithms.
We recommend focusing on two questions in the short-term:
· How are physicians treating patients now? Using anonymized cross-therapy area patient level data (APLD), we can begin to understand how physicians stratify patients and which treatments they select. This goes well beyond Rx behaviors by analyzing comorbidity profiles, patient events, polytherapy and regimen progressions.
· Why are physicians selecting specific products and regimens? Using multiple data sources, we can project market research insights to the broader physician universe. This differs from segmentation — we are not projecting segments based on groups of perceptions or attitudes. Rather, we are projecting individual perceptions and attitudes that can be combined with the patient treatment protocols to customize messaging.
Together, these two elements will help us understand:
1. Where physicians are on their adoption journeys
2. Where their patients are on their respective treatment journeys
3. How physicians determine treatment choices for each patient
Over time, we expect manufacturers to add new data sources that improve nano-segmentation.
What data do we need to project market research insights?
At a minimum, we must leverage a combination of Rx and APLD to isolate key decisions that reflect the behaviors and attitudes of interest. However, this requires appropriate analogs within the market of interest. When these analogs don’t exist, we will expand our search to include other relevant therapy areas. ZS has developed a data platform called ZAIDYN Panorama which does exactly this, prescribing behaviors across multiple therapy areas to draw insights on treatment decisions.
How will nano-segments help improve omnichannel marketing?
Marketers already use the concepts of the patient treatment journey and the healthcare provider adoption journey. However, we only apply these tools at a segment or national level. This allows marketers to develop positioning and messaging appropriate for the target universe in aggregate, but not necessarily for any one customer.
This approach takes these tools and applies them at a much more granular level, asking:
· Where is an individual customer in their adoption journey?
· What are the most common next steps for like customers who were in a similar position recently?
· Which patients should we focus on for a particular physician?
· What perceptions does the physician have around the treatment options for these patients?
· What message do we want to deliver to this physician?
· How should these messages evolve over the planning cycle?
Clearly, we cannot apply this approach for every customer right away. However, these tools will give us a framework to develop the right messages, assign the right messages to individual customers over time and flag specific cases that don’t pass the red-faced test.
