Next Best Action: High-performing operating model

Mark Davies,

This is the third blog in a series that dives deeper into adopting a next-best-action approach to customer engagement.

My first two blogs discuss why enterprises should embrace a one-to-one approach and how to get organizational alignment around your one-to-one vision. Today we’ll dive into the people, processes, and organization essential to efficiently take advantage of AI powered next-best-action (NBA).

Let’s start with describing the governance that organizations should use to control business-as-usual activity.

I’ll explain how to make sure that critical stakeholders are fully engaged and that the interests from different lines of business are effectively balanced to maximize business outcomes and customer experience.

This requires a few key concepts to work well:

  1. First, it requires a single, centralized governance event that will manage the intake and prioritization of the work. We call this the Execution Forum. They receive new marketing briefs or ideas from all members of the stakeholder community across all lines of business.

  2. Second, this prioritization must be run based on agreed rules. Usually, work is prioritized based on the business value that it is expected to drive.

  3. Third, the prioritization process must be transparent to all of the relevant stakeholders. It’s critical to have established, transparent, and agreed-upon rules so that every line of business feels like they have a fair chance of getting their NBAs into market.

  4. And fourth, the work needs to be owned by the right teams – the business changes managed through this process need to be owned by business resources. Pega is designed to support this, enabling business teams to deliver at the speeds dictated by business conditions, while technical teams can deliver through a separate process at a more cautious speed to mitigate technical risk.

Next, I want to describe the process that organizations should use to manage business-as-usual activity; how they feed the NBA machine with new offers, actions, communications, or messages.

The best performing organizations have put together agile structures and processes that allow them to take in demand, effectively prioritize that demand (the new offers, actions, communications, or messages), and then ensure that they get implemented. Some use a scrum approach and some use a Kanban approach, but all follow a similar process.

This process typically consists of three stages:

  1. Manage the intake of new requirements, offers, propositions, or messages or feedback from the live environment; evaluate that demand; and then prioritize based on forecast business value. We’ve described above how critical it is to ensure that this is effectively managed.

  2. Turn the new requirements into real systems assets that can be used within the NBA prioritization framework, pulling together creative, defining the business rules that restrict who can receive particular NBAs based on eligibility and suitability, and creating the AI models that will determine which customers are likely to want which offers. Then – and this is really important – running a simulation activity across a subset of the customer population to see how NBAs are going to be distributed. This is to give you confidence that the NBA mix is in-line with expectations.

  3. The final phase is deployment into the runtime environment. Now the machine takes over and works out when to recommend the new NBAs that you have just created, balancing customer interest and company objectives to work out the right offer for each customer. In this phase your team will be focussed on monitoring and reporting, making sure that the machine is continuing to work as expected, feeding back the performance of the NBAs into the next phase.

Finally I want to describe the teams that you will need to put in place to support this.

In line with the agile approach that we’ve described above we recommend that these are self-contained, self-managing teams, containing all of the skills necessary to deliver business requirements.

The core team will often begin as key resources within the implementation team, gain experience through the initial implementation, and then become the backbone of your business as usual capability. They’ll manage everything from the simplest requests like the addition of new actions, models, or rules, to more significant changes, like redefining an arbitration formula.

There are five role types permanently assigned to this team:

  • Team Leader: Ensures the business benefits of next-best-actions are delivered for their business area. Manages the day-to-day activities required to support effective performance of the scrum team.

  • Lead Decisioning Architect: The most senior Pega Marketing and Decisioning resource in the organization and the hands-on leader of the NBA Execution Team, defining and driving implementation of standards, strategy architecture, and arbitration frameworks.

  • NBA Designer: Works with Lead Decisioning Architect on the implementation – specifying, designing, and configuring one-to-one engagement strategies.

  • NBA Analyst: Assists business stakeholders with formalizing / documenting their requests, and helps them get their ideas, products, and propositions into market more quickly. Think of them as a business analyst that has deep understanding of how decisioning should be used and applies that expertise to shape the incoming demand so that it takes full advantage of Pega’s AI and analytics.

  • QA Tester: Manages the testing process, ensuring that any changes are ready to be simulated.

In most organizations there are other resources seconded into the team, including a business transformation consultant that will ensure that the organization continues to put in place the right training, incentives, and organizational changes; a data scientist/analyst that locates the data and external models required in the decision strategies; and a lead system architect that will make sure the necessary infrastructure and integrations are in place.

When the volume of new requirements for the team increases, we’ve found that it is more efficient and cost-effective to add less-experienced resources, focused delivering the simplest “business-as-usual” changes, such as the addition of simple next-best-actions, straight-forward rule changes, or addition of new adaptive models. This team is comprised of NBA analysts and data resources mentioned above, as well as another resource – the NBA specialist. An NBA specialist is the least experienced configuration resource, set up to deal with a large volume of straightforward work.

Structuring the teams in this way, using a central group expert in one technology, across all channels and all business objectives, results in improved customer experience, higher throughput, and more flexibility.

 

Learn more:

Tags

  • Challenge: 1-to-1 Marketing
  • Product Area: Marketing
  • Product Area: Pega Marketing
  • Topic: Decision Management

About the Author

In his position as Pega’s Senior Director for Business Excellence, Mark Davies helps leading global organizations apply AI in innovate ways to engage more effectively with customers, increase revenue, and drive operational transformation.