Commercial outcomes are significantly amplified by orchestrating a cohesive experience for customers across channels. Artificial intelligence innovations such as ZS’s VERSO™ Orchestration Engine—a cloud-based solution that provides “next best actions” for sales and marketing by improving the quality and timeliness of customer engagements—helps shape that deliberate customer experience and related business performance.


Top-10 global pharmaceutical company Merck partnered with ZS to leverage the Orchestration Engine for a comprehensive field suggestions program, adding a competitive advantage for its commercial organization. We sat down with Deirdre Milici, who leads Merck’s U.S. adaptive customer engagement strategy, and ZS principal Saby Mitra, a leader in ZS’s customer-centric marketing practice, to learn more about how Merck’s field suggestions program has transformed the company’s sales and marketing practices through a holistic approach, including advanced analytics, AI and technology.

Deirdre Milici: We saw the opportunity to leverage artificial intelligence to improve both the customers’ engagement and their experience with us. By having deep customer insights and suggested actions, we expected our conversations to improve with HCPs so they were more relevant and more effective. In addition, we saw it as a way to increase our execution of strategy and field efficiency.


Saby Mitra: Based on ZS’s primary research, we know that two-thirds of the drivers shaping a positive customer experience are well within the control of the commercial organization, including their sales teams. This program is all about improving customer impact through the integration of field and marketing, and surfacing high-value customer insights that are difficult to uncover using traditional techniques. This also created an opportunity for Merck to advance a more analytics-driven culture and a “winning with data” mindset.

DM: Our representatives do a good job of analyzing their business with the data they have today, but they didn’t always have everything that [our team] could make available through the suggestion engine (Orchestration Engine). The suggestion engine has been wonderful in helping us combine a variety of data sources to come up with insights that help the representative. It provides more information for them to prepare for the customer interaction.


The other aspect I think is important is that we’re able to move very quickly with the data and update it frequently. As the data changes, the suggestions can change, and we don’t have to wait for a planned release. With machine learning, we learn more about what our representatives are acting upon and what’s important to our customers, allowing the suggestions to be more relevant as we utilize the engine more and more. 


SM: The ability to create very personalized customer engagements is one of the key opportunities that data science is providing in this program. Additionally, data science techniques are helping to drive the timeliness of the triggers and enabling reps to prioritize where to put their energy. The machine learning is also based on crowdsourcing and how reps are engaging with different types of suggestions. That learning helps optimize the next set of suggestions that are served back to the field.

DM: Aligning the organization around the value proposition is critical. Everyone in the company—sales, marketing, analytics, operations—who’s touching this capability needs to understand the value that it brings to the organization and our customers. Why are we engaging with the suggestion engine? What does it mean for our customers? What does it mean for the field, and what does it mean for the overall business? Before that value proposition is understood, you can’t move forward.


Once we began building out the insights and the suggestions, it was clear that content was king. We had to ensure that the use cases or insights were spot on: aligned to marketing priorities and valued by the field. When we involved marketing and sales to inform the use cases, we were able to deliver a high-quality product. Then the adaptive customer engagement team partnered with ZS to propose new ideas and business rules, and to shape the insights and suggestions based on strong knowledge of artificial intelligence and the business goals. 


The final piece is sponsorship. Strong sales and marketing leader sponsorship provides that excitement to the field about how they’re going to leverage technology with the goal of improving customer engagement. Change can be hard, especially when representatives have managed without AI in the past and have strong relationships with their customers. Once the field learns how suggestions may improve their discussions with their customers, it becomes clear that this is a great way to go. With good communication and a clear training program, the field will be ready on day one.


SM: One of the critical success factors was having the blend of analytical sophistication with deep domain expertise. AI can analyze historical data and provide great personalization at scale, but transparency may get a little compromised: “I don’t exactly know how you came up with this type of suggestion or this type of recommendation.” When rules cannot be extracted from data, we infuse brand and domain expertise into the suggestion design. That was a win-win situation that drove the value of the program. Additionally, the quality of the suggestion business scenarios, the availability of a healthy data mix, and the speed of computing using Amazon Web Services (AWS) contributed to the success of the program.

DM: Start with the value proposition and the “why” behind suggestions to socialize it across the organization. Ask senior leaders to sponsor the program, and identify change agents that will champion the suggestion engine. Then assess and understand the level of knowledge around digital, artificial intelligence and machine learning with all the stakeholders and end users. Develop a plan to communicate consistently across the organization, aligning to company goals, and help people understand how the suggestions engine can help them meet their goals. Start early and communicate often. 


Once the field is using the engine, plan regular feedback sessions with them and incorporate their ideas to continuously improve. This allows the field to contribute to the content of the engine and creates more buy-in. Monitoring utilization and evaluating the suggestions using the analytics portal and surveys is also critical to ensure a quality product. Finally, be sure to measure your success with specific metrics, sharing the progress and success widely.


SM: One change management aspect that was important for Merck was establishing a robust measurement strategy to assess the program’s success and reassure the business of the value of the program. Towards the tail end of the pilot period, we instituted a field survey. We tried to collect data, understand the field sentiments and optimize suggestions accordingly. Then we analyzed secondary data to confirm a direct correlation between suggestion usage and business performance. It was a great opportunity for us to build a story around the impact from the technology and to use that as a vehicle for other franchises that might be skeptical about the change.

DM: The feedback and actions from the field have told us that this is a success because they are utilizing it regularly. [The suggestions program] really helps them focus on the priorities. The suggested action becomes the pre-call plan and they have the option of acting upon it or dismissing it. In some cases, it has led to more time with customers because they’re not trying to gather information. They’re able to get right to the business at hand.


They were also able to recognize new opportunities that they hadn’t seen before because we coupled data together and shared new data sources or insights with them. They found it very helpful not only in their discussions but also to expand interactions across channels. They sent more relevant emails and were able to improve their call plan adherence. The bottom line is always, “Were we able to improve our business results?” And the answer is yes. When our representatives completed the suggested action, we saw better sales performance with those customers. 


SM: We were looking into both leading and lagging indicators to assess the efficacy of the program. While it took a little time for the lagging indicators such as sales performance to show a statistically significant change, we were able to measure leading indicators such as customer engagement rates much earlier in the program to understand directional improvements. 


The other key aspect that we were trying to establish was the sustainability of business results. If we run the measurement over a period of, let’s say, six months, do we sustain the benefits? We observed that we did, and that was very exciting to Merck leadership.

DM: We expect suggestions to expand across the organization. It really is a key capability that supports our digital transformation and will help us continue to drive business results as well. It’s a nice marriage between what we do in the in-person channel and the digital channel. We are going to look for new opportunities to expand the use of our data analytics and insights, automation and engine learning to enable that seamless, integrated customer engagement across channels.


SM: We initially focused on optimizing field engagement and have scaled the program across multiple franchises as part of the current program. We are continuing to analyze data to estimate the impact of optimizing other channels. My sense is that the program, from the success that we have shown in the U.S., has already generated excitement outside of the U.S. to take advantage of the learnings.


DM: As the pharmaceutical industry is looking at how we can engage our customers across digital channels, I think the suggestion engine will really help to take us to that next level.