Our experts share the essentials of maximizing rep engagement with AI-enabled decision support programs. Learn how to:
- Make stakeholders co-authors
- Identify and market the right metrics
- Build a practical foundation that can scale well
- Incorporate the human element into AI
Thanks everybody for taking time out of your day to join us for today’s webinar on the Essentials for Maximizing Engagement in AI Decision Support Programs.
I’d like to introduce today’s presenters. We have with us James Wong who is our Vice President of Product Management and working with our customers on over a hundred brand deployments. James has really seen the diversity of questions life science companies are asking of AI decisions support programs as well as the operational obstacles and hurdles to achieving success.
We’re also joined by Jina Corneau who is our Product Manager in charge of Data and Reporting. Jina works with our customers to really distill their goals and develop metrics and reporting for every stage of their program.
So with that, I’d like to pass the presentation over to James.
Great. Thanks Yigal.
We have seen that AI enabled decision support has become such an important tool in connecting a brand strategy with effective execution. From the case studies that we’ve seen with our customers, we see meaningful results ranging from increased CRM satisfaction or through to higher HCP engagement because you see more compliant email open rates. Ultimately, this leads to financial impact for the brands in the form of increased sales lift.
However, for us to get to these results, we need to understand how we start with effective initial engagement, sustaining this engagement, and ultimately, how you scale out these programs so you can achieve these results.
Thank you so much once again for investing in the time to be with us today. Even though the organizational context you come from varies significantly, we do hope that there are a few thought starters that we can provide today as you consider how you implement or expand decision support at your company.
To begin, I’d like to provide a roadmap of the areas that we plan to cover with you today. To kick things off, Jina will start by laying out a model for how we can empower stakeholders, in particular Sales and Marketing, how they can collaborate together to both launch and maintain decision support. At this point, we’ll also highlight the role of other stakeholders such as field managers in coaching for effective engagement.
Then Jina will demonstrate how we progress from metrics that are operational leading indicators that give you an early heads up on how things are working out to the more strategic metrics that might help you analyze or assess financial impact. We’ll also discuss the role of metrics in how you can drive a continuous program around these metrics.
With that in place, let’s also consider how you establish a strong proving ground and how you go about building on that foundation for impact so you can build a case for scaling the program across your organization.
And then finally, we present a few ways of how you can practically apply AI to leave room for human involvement so that you can further drive engagement with both headquarter’s Marketing teams as well as driving dynamically adjusting content for your field teams.
I’ll now hand over to Jina to take us through the first half of our content.
Great. Thank you James. Well as James mentioned, engagement is driven by four essentials. And the first to stakeholders. There are many stakeholders in the launch of any new project from executive sponsorship to IT. But what I’d like to focus on right now is Sales and Marketing.
I believe most of us can agree that a collaboration between Sales and Marketing is desirable to the success of any campaign. But what we have observed time and again, is that the most successful projects involve Sales from the very beginning. There are several key stages in the project launch process where Marketing and Sales collaborate as seen here. And it all starts with project Setup and Design. Sales has historically been removed from the strategy phase, but we have seen that the projects that have enjoyed the highest engagement are those that have included sales reps as co-authors. By including reps at the start, you’re able to better understand the end-to-end rep workflow, the information that reps find meaningful to make decisions, and the points at which rep feedback is the most effective.
During the UAT and testing phase, reps are receiving real data and are able to provide feedback creating an early, realtime loop between Sales and Marketing for process refinement.
As we move on during Hypercare and into Steady State, those feedback loops are further refined and grown to ensure that campaigns are translated from the design table to execution as expected and that they’re running as effectively as possible. An example of collaborative refinement at this stage includes the adjustment of active feedback channels to collect the most important information back for Marketing.
Early and continuous collaboration between Sales and Marketing includes really both logistical and psychological advantages. Logistically, we’re able to address design and process concerns at flexible stages. And we’re building a product that reps will actually use. Psychologically speaking, reps who feel that they’re co-authors really have a lot more buy in, often serving as project cheerleaders and promoting the project to others.
All of our observations are taken from the field and also from customers in basic design principles. And we’ve seen that the best way to really ensure your project is best positioned to be highly engaged is by including end users from the start.
Building on this idea of early and continuous collaboration between Sales and Marketing, feedback channels are really the key to success here. There are a variety of ways in which Marketing communicates with Sales in terms of campaigns, customer journeys from adding HCPs to setting particular logic rules to trigger response activity in the reps. But an effectively designed project also ensures that there are a variety of ways in which Sales can communicate back to Marketing. One of these feedback loops involves dismissal surveys. These surveys are the reps’ opportunity to provide specific feedback to Marketing on why they will not action a particular suggestion they received.
The tendency here might be to assume that if a suggestion was dismissed, it was ineffective. But these surveys actually allow reps to be more specific. For example, a rep may believe that a particular channel is not appropriate for an HCP or there may be a specific short term obstacle to engaging that HCP. This feedback can then be collected and used to adjust say that HCP’s inclusion in a campaign or journey.
The last thing I’d like to talk about in terms of empowering stakeholders is the vital role that sales managers play. Successful projects not only create feedback loops between Sales and Marketing but within the sales teams as well. Again historically, sales managers have been a little removed from the strategy loop and have had limited visibility into their team’s promotion of campaigns and actions. But now, sales managers have greater visibility into both through insights reports.
While insights reports do provide engagement and synchronization metrics as seen here, they also provide sales managers more detailed information on what their reps are doing from the suggestions and by extension, the campaigns that reps engage or avoid most, to the information and data points that reps utilize and find most useful prior to a call. A sales manager can now see more contacts to run engagement metrics which better equips them for more effective coaching during their one-on-ones. For example, a District Manager in a district where rep engagement begins to drop, a District Manager can dig in and see exactly which reps’ habits have changed and formulate a constructive conversation for the next ride along.
In one instance where we saw this happen, quick turnover had led to lower engagement in the district as the new reps didn’t understand the project or what the suggestions were, how to use them. District Managers were able to identify these reps and provide one-on-one coaching to not only explain how to utilize the suggestions, but why they were important by pointing to specific metrics and campaigns.
Sales Managers already receive a plethora of data and information so our insights reports are specifically designed to surface the information and metrics that will best equipment them to have productive coaching and effective campaign promotion. This is achieved through both data-driven charts, qualitative information, contacts, and curated insights as seen here. All of this ensures that the feedback that the reps are providing and their own feedback channels is accurate and reflective of what’s happening in the field. And by ensuring that that information is accurate as possible, we know that what Marketing is receiving can be better used to adjust their campaigns and strategies.
Now that we’ve discussed the stakeholders, I’d like to focus on the metrics that those stakeholder use to measure and communicate or market success.
We divide metrics into three basic categories. We have Engagement or what you can observe happening within the system and how people are using the system. We have Execution or what’s actually happening in the real world. What are reps doing and how are HCPs responding? And finally, we have Impact or what meaningful business or financial impact does all of this have? These metrics range from tactical to strategic and we typically see an evolution of these metrics over time.
While we’re all interested in the business impact of a new project, it typically takes about six to nine months to see the full impact. In the mean time, there are a variety of metrics that can be measured as soon as one to two weeks following deployment that are just as critical to success.
Here you can see a timeline of the progression of the metrics that are important to a project. The metrics that you track will evolve over time. And the first that you’ll be most focused on is engagement. Engagement is the foundation and until you have success here, it’ll be pretty hard to advance to the next stages. Engagement starts with Platform Performance. So here we’re measuring how is the platform actually delivering information? How are the frequency of suggestions received? Are they going to the right reps? Is it the right channel mix? From there, we can take a look at Field Engagement. Are the reps accepting, dismissing, engaging these? How are they actually even using the system?
Once we have a good handle on Engagement, we move on to measuring Execution. This starts with Field Behavior measuring rates of activity by channel, content, customer coverage, messaging, et cetera. And while there are a variety of standard metrics available at this and really at every stage, there’s also a lot of flexibility for customer defined metrics. For example, at this stage, a customer may want to measure engagement and activity down to the message level. Perhaps there’s product that has multiple indications and a balance on the messaging on all of those indications is really key to this particular customer. That’s something that can be done at this level.
Going on, there’s HCP Engagement where you can measure things like interactions by channel, channel affinity, message affinity, et cetera. And we’ve seen this one actually be of particular interest to customers when channel mix is a high priority. We’ve seen a lot of great success in defining and tracking custom metrics communicate that through the field.
Following success with Execution, you can finally move on to the metrics that demonstrate Business Impact. We have HCP Behavior which can be anything from measuring customer journey transition to event attendance to open and click through rates on emails. And then Business Performance. What are the total and new prescriptions? What are sales? What are new patients, patient switches, indications and more? Of course the ultimate goal in all of this is changing HCP behaviors to drive business growth with constant tracking throughout the process to provide greater insight into progress and performance for iterative improvements.
The key here is really the evolution through the metrics from tactical to strategic and adjusting your rules and strategies along the way. What’s going to ensure long term adoption and engagement though is then using those metrics to communicate progress and success to all stakeholders, particularly the end users. When reporting is used as a common language, that’s when we see the most success for our customers as it’s much easier to motivate a team and rally around initiative if you’re all speaking the same language.
Building on this idea of reporting as the main mode of communication, it’s also important that your metrics actually lead to action. This is another critical feedback loop that connects strategy to execution and back. We start by curating and moderating the most impactful metrics whether it’s engagement as a suggestion message or even business rule level. For example, let’s say we have configured a particular business rule to create a suggestion for reps when there is low patient conversion. You can then begin to collect feedback about that metric using a variety of feedback channels. Engagement metrics can provide insight into how frequently this particular rule appears and how often reps are engaging with it. And dismissal surveys can provide qualitative insight into why a rep may not be accepting it. Perhaps there is feedback that the message is no longer relevant or that the suggested channel isn’t ideal. But from that feedback, you can then start to learn which rules or even campaigns may need some adjustment. Again, following the example, perhaps based on that dismissal feedback, you might decide that a contact refresh is needed or a different channel is more appropriate and you can reconfigure your rules. Once these changes are made, the next step is to assess the impact of the change and again start monitoring these metrics and move back through the loop.
As mentioned in the example, there are a variety of metrics available to customers from high level down to a very granular level. And one of the really most impactful and widely used reports we’ve seen is the business rule report seen here. This report shows metrics down to that business rule level which is the most granular level. Here you can see highlighted, the business rule that we were talking about, the low patient conversion and you can see the actual engagement metrics tied to that rule. There are a variety of actions that can be used with these metrics to ensure that there is long term engagement and success of these projects.
So with that, I am now going to turn it back over to James who’s going to move on to the next two essentials for maximizing engagement starting with building a practical foundation that ensures scalability.
Great. Thanks Jina. So next up, I want to talk about building a practical foundation to ensure that your program is scalable. Because ultimately this is the key to engagement over the long term. And we’ve observed that without specific business challenge or opportunity to rally business stakeholders around, decision support programs end up getting out of touch and what you end up with something that theoretically works for everything, but really practically works for nothing. So the first question that I’d like to address is where’s a good place to start? Let’s start by evaluating the type of product market or competitive opportunities that decision support can help address.
We could identify a brand that’s relatively mature and is looking to respond to a competitive threat. Or maybe we’re looking to address an upcoming loss of exclusivity event. Next we could consider what teams or therapeutic areas you might want to launch with decision support as you are well aware, how you go to market with oncology products which tends to have more diverse go-to-market team members such as MSLs or nurse educators or reimbursement managers looks very different to a primary care team who has more of an established call plan, or a vaccines team who might engage with not only physicians, but also pharmacists and pharmacies as well.
Next we consider what are the most important priorities that decision support should start focusing on? Perhaps you want to reinforce messaging for certain segments of physicians or address referral drop offs because you’re seeing that sales are being impacted by the conversion between primary care physicians and specialists, not subsequently writing those scripts.
And finally, especially for those of you who are in a global role, what market is most appropriate to launch a decision support program? Perhaps there are markets with new leadership whose focus in upcoming year is in digital excellence or digital related initiatives. Or, are there markets with teams who have demonstrated a culture and willingness to experiment with novel go-to-market technologies?
I want to make it clear that this isn’t about picking an absolutely perfect starting point. Instead, what we really want to provide are some thought starters and dimensions by which you can be explicit about where you might want to start. And most importantly consider why that is a great starting point.
To bring this concept to life a little bit, let’s discuss a more tangible application of this. A common starting point we’ve observed is typically within the mature brands where sales have started to stagnate and field teams might be plateauing on field force satisfaction because they’ve been doing the same stuff for a while. Yet, we believe that there is a large opportunity still due to the patient need as well as the addressable market. Perhaps there are additional challenges arising from the fact that — in this case it’s a primary care team — and they have multiple brands in their bags that they have to juggle across different segments of physicians and it’s not always easy to remember what you should detail in the first position or the second position. So often, evolving complex strategies like this is a great fit for decision support.
Perhaps in order to secure buy in with the field force, you want to start with a relatively simple set of use cases around helping reps pace their visits in order to hit their activity based goals. Another area of quick wins is around coordinated activities and awareness of what physicians are doing not only in the visit channel but perhaps in digital channels such as email sends and coordinating across both of those channels.
Finally, North American is often seen as a great starting geography mostly due to the abundance of data with which to drive decision support programs. However, that’s not to say that decision support isn’t possible in other geographies due to some of these data granularity challenges. We’ve seen that with some thoughtful crafting of good decision support strategies and applying advanced analytics to disaggregate some of this data, we’ve seen many customers drive to great success in Europe, Japan and China.
So you may be wondering why I’m talking about starting simple and establishing proving grounds in a section that’s meant to be about driving scale. And really it comes back to setting a solid foundation. Definitely make sure that the solution you’re implementing is scalable across each of these dimensions. And today we won’t be focusing on what it means to build a scalable technology solution. There are many technical elements around how decision support engines should work to serve the various therapeutic areas and to recognize the go-to-market and HCP differences in the various geographies.
Instead now I’d like to cover some of the softer structural elements that are required to go from this solid foundation to being the new go-to-market model for your company. And there are quite a few categories here so I’ll provide a tangible example briefly through each one of these elements.
To get started, when we talk about Governance Model, this isn’t really only about the right effective executive sponsorship which is important. But it’s also about striking the balance between global governance where you can drive technology and process standardization with local governance who can really apply these global standards and configure it in a way that delivers the most value for the local market needs.
Next we talk about Roll-out Process. And this is really about codifying how each of the deployments within each brand or each geography should be conducted.
Now we’ve touched on this next topic a couple of times already but I want to highlight the importance of providing Ongoing Support for your decision support program. A decision support engine no longer feels smart once it becomes stale. So this ongoing support process really helps ensures that the decision support content remains fresh.
Now Incentive Compensation can be a tricky topic given how much of an impact it has on the field. That said, we also know that incentives are a major part of how field teams think about engaging with decision support. Now to be clear, it doesn’t mean that we’re suggesting that we blindly incentivize positive engagement because this can often have unintended consequences. For example, Jina talked about how one of the more valuable signals we get from field teams is when they use these dismissal surveys and we use these dismissal reasons to then refine the engine, to make the engine smarter. If all we’re doing is incentivizing positive engagement and excluding dismissals as an example, we run the risk of losing this important datapoint to make the engine smarter.
Now occasionally, we see that decision support efforts are often too skewed towards or away from IT. Our advice here is really to establish clarity on what the role of IT is throughout the deployment of your program. In a similar vein, often analytics teams are engaged after the fact and simply charged with the mission of, “Go find impact and help us understand what’s going on.” However, ensuring that the right analytics engagement is occurring upfront will not only help provide a sense of co-authorship, which has been a constant theme in our discussion today, but also ensure that the analytics teams have the appropriate business context during analysis that can help often rigor test the results or the conclusions that you might want to draw.
You can certainly start without having tackled many or all of these elements. However, if you’re looking to scale your program beyond the solid foundation, these are many of the structural elements that should be in place.
Now AI comes with a perception of full machine automation that negates the need for any human involvement and therefore, by definition, any human engagement. But we’ve actually seen the most successful usage of AI be a bit of a blended approach where machines learning is doing what it’s good at but so too are the individuals that are interacting and therefore, engaging with this artificial intelligence. So let me show you what I mean.
To set the scene, there are three ways that we can look at human involvement and the oversight of AI. A fully manual approach, a fully automated approach or somewhere in the middle where we have a blended or assisted AI approach. Really it’s when we have this balance that team members see AI as an asset to their contributions and not a replacement.
Here’s one real world case of how we’ve implemented this principle for marketing users in one of our AI products. As context, what you’re seeing here is a part of our message sequence optimization model that as the name implies, is really about driving a personalized sequence of messages for an individual HCP that optimizes engagements such as open or click rates.
On the screen, what you’re looking at is a visualization of a particular optimized sequence. Now a marketer sees this sequence and is able to manually move the order of the messages due to some specific marketing strategy or event that has taken place in the marketplace. Furthermore, he or she can even specify constraints such as, “This message always having to be in this position,” or setting some sort of sequential constraint or relationship between two messages but not always being in a fixed position. Or perhaps this message is only valid within a specific time window where a promotional message is in effect. And with this example, we’re able to drive marketing engagement with AI by allowing Marketing to easily provide input and have oversight into our AI product.
We also see see AI drive engagement when machine learning incorporates human involvement through questions to really curate and improve suggestions. As context, we always curate a list of the most important suggestions for a rep. So in this case, let’s say that we’ve set it so that the top five suggestions are provided to a rep. However, beneath those top five, there are still valid suggestions that are in line with the strategy but that are otherwise suppressed because we are only limited to five and we don’t want to overwhelm the rep.
However, we can start using AI to overlay some human considerations to further curate these suggestions. Could we predict where a rep might be on given day to prioritize suggestions that are more on the route? Could we assess the HCP availability so we predict when the HCP will be available and match that up with an open slot on the rep’s calendar? Could we look at how reps have historically engaged with suggestions to determine what action a rep is most likely to take that day for that account? Could we apply many of the principles of behavioral economics and nudge the rep to act by adding content that demonstrates the value of accepting that suggestion?
When all of these human elements are incorporated into your decision engine, we can actually determine a different order of suggestions that might be shown to the rep. Some suggestions that were previously suppressed may now be compelling enough to poll above the line to show to the rep. And ultimately, all of this helps drive dynamic content that feels more right to a rep and ultimately drives engagement by a rep.
To recap, here are the four essential considerations that we’ve covered today based on observations that we’ve made over a hundred deployments that we’ve partnered with our customers on over the last few years. And regardless of where you are in your decision support maturity, we hope that you’re able to take away a few of these start thought starters from this webinar and we look forward to supporting you in your journey.