Next best actions with Redpoint

Overview

Personalized marketing engagement is key to building lasting relationships with your customers and delivering value across their journeys. To do this, you need to deliver the right message, through the right channel, with the right offer, at the right time. That requires a deep understanding of customer preferences, behaviors, and needs, along with the ability to act on that insight in real time.

Deciding on the best action to take for each customer and each situation isn't trivial. It means balancing business objectives, customer expectations, channel availability, budget constraints, and competitive pressure. It also requires a flexible, adaptive approach that responds to changing customer behavior and market conditions, and that learns from previous actions and outcomes.

Next best action (NBA) strategies help you move beyond static campaign schedules toward more relevant, real-time engagement. Instead of guessing what a customer might want next, or applying a one-size-fits-all approach, an NBA strategy helps you identify and deliver the most appropriate action for each individual, based on their current context and behavior.

Organizations take different approaches to determining next best actions. Some rely on explicit, hierarchical business rules that define the logic and criteria for selecting actions based on predefined segments and scenarios. Others use analytical models that apply data and algorithms to predict the optimal action based on customer propensity, value, and risk. Many combine both approaches, using rules to govern and refine the results of the models.

Regardless of approach, you need a platform that lets you design, execute, and optimize next best actions at scale. Redpoint provides a comprehensive solution that supports a range of use cases and approaches for delivering next best actions in support of personalized marketing engagement. With Redpoint, you can:

• Manage and unify customer data from multiple sources to create a single view of the customer.

• Segment and profile customers based on their attributes, behaviors, and interests.

• Define and implement business rules and analytical models for determining next best actions.

• Execute and orchestrate cross-channel campaigns that deliver next best actions in real time.

• Monitor and measure the performance and impact of next best actions and campaigns.

• Learn from feedback and results to continuously optimize next best actions and campaigns.

This article covers how Redpoint's platform helps you deliver the next best actions that drive personalized marketing engagement and customer loyalty, including the core principles for designing, executing, and optimizing an NBA framework.

What is a next best action?

A next best action is the optimal decision you can make at any given moment to engage with a customer or prospect: the most relevant, timely, and personalized point of engagement that maximizes the customer's value and satisfaction. The definition can vary depending on your organization's goals and strategy. A next best action can also refer to the optimal channel, message, or offer delivered as part of the engagement, or any combination of these.

For example:

• A retailer might send an email with a special discount to a customer who abandoned a shopping cart.

• A healthcare provider might remind a patient to refill a medication or schedule a follow-up appointment.

• A bank might call a customer who applied for a loan.

• A media company might suggest a relevant article or video to a reader or viewer.

Determining the next best action for each customer isn't simple. It requires a deep understanding of the customer's behavior, preferences, needs, and context, plus your business objectives and constraints. The next best action also changes over time as the customer interacts with your brand across touchpoints and channels, so you need a dynamic, adaptive solution that continuously learns from data and feedback and delivers next best actions in real time and at scale.

Rethinking "next best action"

An NBA strategy starts with a simple but fundamental shift: from brand-first communication to customer-first engagement.

In traditional marketing, you plan campaigns in advance, execute them on a schedule, and measure results after the fact. NBA reverses that flow. Each interaction, whether it's an email open, a site visit, or even silence, becomes a new input that informs the next move.

The goal isn't constant activity, but meaningful activity. Sometimes the best next action is to hold back and wait for a better signal. This approach helps you maintain relevance, respect customer attention, and prevent message fatigue.

Redpoint's platform

Redpoint's platform uses artificial intelligence (AI) and machine learning (ML) to analyze customer data from multiple sources, generate insights and recommendations, and orchestrate personalized, consistent interactions across channels. You can configure the platform to fit your goals, whether you want to optimize the next best point of engagement, channel, message, or offer, or any combination of these.

To support next best action use cases, Redpoint offers three core applications:

• Redpoint Interaction (RPI): Create and manage audience segments, smart assets, and interactions across channels. Audience segments are dynamic snapshots of customers based on their attributes, behaviors, and preferences. Smart assets are reusable content elements you can personalize and optimize for each customer and channel. Interactions are rules-based or ML-driven actions that deliver the right message, offer, or recommendation to the right customer at the right time and place.

• Redpoint Data Management (RPDM): Integrate, transform, and enrich customer data from sources such as your CRM, web, mobile, social, email, and offline channels. RPDM also provides advanced analytics and ML capabilities, such as propensity scoring and recommender systems, to help you understand customer behavior, predict outcomes, and generate personalized recommendations and offers.

• Redpoint Realtime Decisions (RTD): Define and execute RTD rules and smart assets that deliver the next best action to each customer in any channel. RTD uses cached attribute lists and contexts to optimize the performance and relevance of the next best action. Cached attribute lists are staged sets of data elements, such as customer details, summary data, model scores, or recommended content, that RTD can use during execution. Contexts represent a collection of decisions tied to where someone is engaging from, such as a section of your website, your mobile app, or your call center.

A unified customer profile that merges online and offline data into a single, accurate view forms the foundation for everything these applications do. In practice, the three work together as a closed loop: RPDM aggregates and normalizes your data so every decision is based on current, complete information; RTD (together with RPI's business rules and models) evaluates that data to determine the best next step; and RPI orchestrates execution across every connected channel. Strategy, data, and execution reinforce each other continuously.

Build your NBA framework in three stages

Every organization's implementation looks different, but most NBA programs evolve through three foundational stages.

1. Define the strategy. Start with clarity on the outcomes you want to drive, such as retention, conversion, engagement, or education. Map your customer lifecycle, define touchpoints, and outline the triggers that should prompt an action. Cadence matters: frequency and timing should reflect customer readiness, not internal schedules.

2. Activate and execute. Once you've established the framework, connect it to your operational systems. In Redpoint Interaction, this typically means configuring audience definitions, decisioning rules, and cross-channel orchestration. Each channel, whether it's email, SMS, mobile, or direct mail, should have content and logic ready to respond to customer signals in real time.

3. Monitor and optimize. Continuous measurement keeps the strategy intelligent. Capture explicit preferences (such as opt-ins or channel choices) and implicit behaviors (clicks, conversions, or inactivity), and feed those insights back into your data and decisioning layers. Over time, this feedback loop refines the accuracy and effectiveness of each next best action.

Use business rules to segment customers

One way to personalize messages and offers is to use a set of business rules executed in a hierarchy: a tree-like structure where each level represents a different degree of segmentation or personalization. You can use a hierarchy like this to assign individuals to a segment representing their stage of the customer journey, or to determine which message is most appropriate for them.

Example business rules:

• Is the customer a new visitor or an existing customer?

  • If new, assign them to the "Prospect" segment and show a welcome message or sign-up offer.

  • If existing, move to the next level of the decision tree.

• Has the customer made a purchase in the last 30 days?

  • If yes, assign them to the "Active" segment and show a cross-sell or upsell message or offer based on their purchase history and preferences.

  • If no, move to the next level.

• Has the customer made a purchase in the last 90 days?

  • If yes, assign them to the "Lapsed" segment and show a re-engagement or loyalty message or offer.

  • If no, assign them to the "Inactive" segment and show a win-back or retention message or offer.

You can refine and expand this decision tree by adding more levels, criteria, and conditions, such as customer lifetime value, product affinity, channel preference, recency, frequency, and monetary value. You can also weight and prioritize rules to account for different scenarios and goals. For example, a customer in the "Active" segment with a low lifetime value might receive a different offer than an "Active" customer with a high lifetime value.

Using business rules in a hierarchy lets you segment customers by behavior, needs, and value, and deliver personalized next best actions that increase conversion, retention, and loyalty. Redpoint's platform lets you create, manage, and execute business rules in a user-friendly interface, and to test and optimize their effectiveness using advanced analytics and ML. You can also leverage cached attribute lists and contexts to keep the next best action relevant, timely, and consistent across channels.

This hierarchical approach is implemented as an audience within Redpoint Interaction. The audience supports the execution of complex selection, suppression, and segmentation criteria in a top-down cadence, enforcing prioritization of the logic. When executed, the audience assigns each targeted individual to one or more segments they qualify for, tagging their record with the segment metadata that represents the offer, message, or campaign they qualify for. You can run this audience as an audience snapshot, which re-evaluates each individual's segment membership at a set interval (daily or hourly, for example) or on demand when triggered by an external process.

Use analytical models to predict customer behavior

Analytical models can also drive next best actions by predicting customer behavior, preferences, and needs. These models draw on transactional, demographic, behavioral, or contextual data, and use techniques such as propensity scores or recommendations. Propensity scores estimate the likelihood that a customer will take a specific action, such as buying a product, renewing a subscription, or churning. Recommendations suggest relevant products or services based on a customer's past purchases, browsing history, or similarity to other customers. Both help you identify the best offer, message, or channel for each customer and optimize the customer journey.

Redpoint's platform supports building, maintaining, and executing analytical models within its own environment, as well as integrating external models from third-party applications. This bring-your-own-model (BYOM) approach gives you the flexibility to use the models and tools you already have. You can bring in external models by ingesting externally generated scores, or by calling your own modeling environment's API. Redpoint's platform also provides a model management interface for configuring, organizing, and deploying predictive models, including models built in your own environment, and lets you access model scores and outputs via REST APIs. You can use analytical models as part of campaign or audience execution, and apply filters, exclusions, and prioritization rules to refine the next best action.

Start with simpler decision rules and layer in AI-driven decisioning as your program matures. Combining analytical models with business rules lets you use the power of data and analytics while keeping the results governed and predictable.

Deliver next best actions in real time

Delivering next best actions in real time is a key challenge, because customers expect personalized, relevant offers and recommendations that match their current intent and context, whether they're browsing your website, using your mobile app, calling your contact center, or visiting your store.

Redpoint's RTD (Realtime Decisions) services are microservices you invoke via REST APIs. They can retrieve staged next best actions or recommendations, execute rules or models, apply filters or exclusions, and return responses in JSON format. You can configure and deploy RTD services according to your business requirements and use cases, and scale them across environments.

Using RTD services, you can choose between two approaches to deliver next best actions in real time:

• Use next best actions that are pre-computed and stored in Redpoint's Customer Data Platform (CDP) database, exposed to RTD as a cached attribute list.

• Dynamically generate next best actions based on the latest data and logic. This may involve integrating with external modeling services, such as AWS SageMaker or Google Cloud AI Platform, to execute analytical models that produce propensity scores or recommendations on the fly.

Both approaches help you deliver the most relevant, optimal next best action for each customer, in real time, across any channel or touchpoint.

Because customers interact through multiple channels, real-time delivery works best when every channel is coordinated. Map each journey across email, mobile, web, and offline channels; set cadence and frequency rules to prevent over-communication; and keep messaging consistent, personalized, and optimized for each delivery method. An SMS isn't formatted like an email.

Common implementation challenges

Implementing an NBA strategy at scale tends to surface a few recurring challenges:

• Content scalability: Personalization at scale needs a steady flow of modular, reusable content. Start with key journeys or segments, then expand incrementally as your framework matures.

• Cross-team coordination: NBA relies on collaboration among data engineers, analysts, creatives, and campaign operators. Shared key performance indicators (KPIs) and clear ownership for each process step help maintain alignment.

• Goal definition and measurement: Decide early what success looks like, whether that's an increase in engagement, conversion, retention, or satisfaction, and make sure your metrics tie back to those objectives.

Plan your organizational strategy

Before implementing next best actions in real time, define an organizational strategy and pattern that aligns with your business objectives and customer needs. That means making several decisions and trade-offs:

• What level of granularity and personalization do you need? Next best actions can be defined at the segment, persona, individual, or even micro-moment level. Granularity determines how tailored and relevant the next best action is for each customer, as well as how complex and costly the implementation becomes.

• Which channels and touchpoints will you use? Next best actions can be delivered across email, SMS, web, app, call center, or store. Your choice depends on customer preference, channel effectiveness, and channel availability.

• What cadence and frequency work best? Next best actions can run on a batch or real-time cadence, depending on the channel and customer journey stage. Cadence affects how timely and fresh the next best action feels, as well as customer response and fatigue.

• What content and format will engage and persuade? Next best actions can include offers, recommendations, messages, images, videos, or interactive elements. Your choice depends on the customer profile, channel characteristics, and creative design.

• Will you use business rules, models, or both? Business rules can define eligibility, priority, or exclusions for next best actions, while models can supply propensity scores or recommendations based on historical or real-time data.

Defining this strategy up front helps you deliver consistent, effective customer engagement across every channel and touchpoint, and gives you a baseline to test, measure, and refine over time.

Use AI to accelerate NBA programs

AI enhances NBA programs by automating repetitive tasks and improving precision:

• Content support: Generative AI can help create or adapt messaging variations to keep content fresh and reduce creative bottlenecks.

• Decision refinement: ML models can improve decision accuracy by learning from historical engagement data.

• Rule simplification: Natural-language interfaces make it easier for business users to build or adjust decision rules without deep technical knowledge.

AI works best layered on top of a strong foundation of strategy and clean data. It accelerates your NBA program but doesn't replace the core work of understanding your customers and executing with discipline.

Practical applications by industry

NBA concepts extend across industries and use cases. In every case, the value lies in anticipating customer needs and engaging with contextual relevance.

Healthcare: Recommend next steps in a care plan, such as appointment reminders or educational materials.

• Preventive care recommendations: Use Redpoint's real-time segmentation and identity resolution to trigger personalized outreach (email, SMS, portal notification) for screenings or wellness visits based on patient profiles and historical data.

• Medication adherence nudges: Integrate pharmacy and claims data to identify non-adherence, then use Redpoint's journey orchestration to deliver tailored reminders or incentives across preferred channels.

• Post-discharge follow-up: Detect discharge events through integrated electronic health record (EHR) feeds, then trigger NBA workflows recommending follow-up care, surveys, or home health services, personalized by condition and demographics.

• Chronic condition management: Combine wearable data, claims, and engagement history to deliver personalized content and coaching through Redpoint's omnichannel engine, such as app push, email, or call center.

Financial services: Provide proactive fraud alerts, loan-application follow-up, or guidance for major life decisions.

• Loan application follow-up: Trigger a personal outreach, such as a call, for a customer who has applied for a loan, so a time-sensitive decision doesn't stall.

• Smart financial planning advice: Use Redpoint's behavioral analytics to identify life stage or financial goals, then trigger NBA content (budgeting tips, investment options) through web personalization or advisor outreach.

• Fraud prevention alerts: Detect anomalies in real time using Redpoint's streaming data capabilities, then trigger NBA actions like account locking, verification prompts, or security tips through SMS or app.

• Credit health improvement tips: Segment users by credit behavior and deliver NBA suggestions, such as paying down specific debts or opening a secured card, through personalized dashboards or email campaigns.

Retail: Deliver personalized offers based on purchase history and browsing behavior.

• Hyper-personalized product recommendations: Use Redpoint's AI-driven propensity models and real-time web behavior to suggest products dynamically on-site or through retargeting ads and emails.

• Post-purchase cross-sell: Show a cross-sell recommendation immediately after a customer completes a purchase, personalized to what they just bought.

• Loyalty program engagement: Identify dormant or high-potential loyalty members and trigger NBA actions like bonus point offers, tier-upgrade paths, or exclusive events through preferred channels.

• Cart abandonment recovery: Detect abandonment in real time and trigger NBA workflows offering incentives, alternative products, or urgency messaging through email, SMS, or app notifications.

Media and publishing: Recommend content and grow subscriptions based on consumption behavior.

• Content recommendations: Suggest a relevant article or video to a reader or viewer based on what they've already consumed.

• Subscription and trial upsell: Offer a free trial or a subscription upgrade to a reader or viewer who has consumed a certain amount of content.

Example NBA process

The following process illustrates how data and decisioning come together to operationalize an NBA strategy:

1. Data inputs: All available customer and prospect data feeds into the NBA environment, including transactional history, behavioral engagement (website, mobile, in-store), and external analytics. The richer the dataset, the more refined and personalized the resulting actions.

2. Unified customer profiles: Data management systems consolidate and clean this information into comprehensive customer profiles (Golden Records). These profiles are the foundation for all subsequent segmentation and decisioning.

3. Segmentation and rule-based decisions: Decisioning happens in multiple phases: high-level segmentation, eligibility rules, and prioritization. Business rules can determine, for example, which customers shouldn't receive certain campaigns based on recent activity. Business users can execute these decisions without heavy technical support.

4. Advanced decisioning and staging: NBA decisions can be enhanced with analytical models and ML, with results staged and ready for execution at the right point in the customer journey.

5. Execution across channels: The system delivers communications across multiple channels, such as email, mobile, web, and more, at the right time and in the right context. Real-time decisions and batch-triggered communications stay synchronized to ensure consistent messaging.

6. Continuous updates and optimization: NBA outputs are recalculated regularly based on incoming data and user interactions. For example, a recent in-store purchase could immediately update the next best action from a promotional offer to a follow-up satisfaction survey, so the system always reflects the most current, relevant action for each individual.

Get started: practical resources

Use the following resources to move from strategy to execution.

1. Unified customer profiles

Why it matters: NBA relies on accurate, real-time data about each customer.

How to get started: Use Redpoint Data Management to merge offline and online data into a single, clean profile. Focus on identity resolution, deduplication, and making sure each profile reflects current behaviors and preferences.

Related resources:

• Redpoint Data Management overview

• Best practices: hygiene, matching, and identity resolution

2. Segmentation and decisioning

Why it matters: Determining the right next best action requires knowing which group or persona the customer fits into at any given time.

How to get started:

• Build segments using behavior, lifecycle stage, or explicit preferences.

• Layer in business rules for eligibility and prioritization.

• Test simpler rules before scaling to more complex, AI-driven decisioning.

Related resources:

• About RPI's standard selection rules and basic selection rules

• Create a standard selection rule in RPI

• Build a segment in Redpoint CDP

3. Cross-channel orchestration

Why it matters: Customers interact through multiple channels; NBA works best when every channel is coordinated.

How to get started:

• Map each journey across email, mobile, web, and offline channels.

• Set cadence and frequency rules to prevent over-communication.

• Keep messaging consistent, personalized, and context-aware, and optimize it for each delivery method, such as email versus SMS.

Related resources:

• Best practices: campaign prioritization

4. Monitoring, feedback, and optimization

Why it matters: NBA is a cycle, not a one-time deployment. Continuous learning improves accuracy and relevance.

How to get started:

• Capture engagement metrics and customer responses.

• Feed results back into the unified profile and segmentation layers.

• Use A/B testing and iterative analysis to refine content and rules.

Related resources:

• Getting started with the RPI reporting hub

5. AI and automation opportunities

Where it helps:

• Content generation (emails, subject lines, web personalization).

• Real-time decisioning and prioritization.

• Natural-language rule creation for faster deployment.

Related resources:

• Using generative AI to create content

NBA quick-start checklist

Use this checklist to get your NBA program up and running efficiently. It's a practical, step-by-step guide to move from strategy to execution.

1. Define your strategy

2. Prepare your data

3. Segment and decide

4. Orchestrate and activate

5. Monitor and optimize

6. Leverage AI and automation

Conclusion

Redpoint provides a flexible solution for delivering next best actions in real time, and that flexibility shows up in a few key ways. Next best actions can…

• Operate at multiple levels and mean different things to different organizations, depending on their goals and strategies.

• Run on a batch or real-time cadence, depending on the channel and customer journey stage.

• Be based on business rules, models, or a combination of both, depending on the data and logic available.

An NBA strategy isn't a one-time implementation. It's a cycle of learning and adaptation. Define your approach, act on it, measure the outcomes, and adjust. Over time, that loop builds a self-optimizing system that grows more accurate, responsive, and efficient.

Redpoint supports this flexible implementation model end to end: you can configure and use RTD services, integrate external modeling services, and deliver the most relevant, optimal next best action for each customer, in real time, across any touchpoint, turning next best action from a series of one-off campaigns into a continuous, adaptive conversation with your customers.