14 6 月, 2017/Aktana/CRM Suggestions

How does Aktana manage to address the needs of a diverse customer base in disparate markets—without building custom products? We’re sitting down with Bob Flint, the principal engineer for Aktana’s Decision Support Engine (DSE), to discuss the importance of configurability and explore how one flexible platform can address the needs of pharmaceutical companies around the world.

Overall, why is configurability so important in decision support design?

Different Pharmas use CRMs differently, adhere to different regulations and follow different strategies and practices in terms of guiding their sales forces. Even within a single rep team at one pharmaceutical company, there’s high variability in the behavior of individual reps.

This is definitely NOT a situation where an inflexible, “one-size-fits-all” product offering will be viable. We’ve been working on our current product set for more than seven years now. In that time, we’ve gone through three major redesigns of our product so that it reflects all that we have learned about the variety of pharma practices.

What are some ways pharmaceutical practices can vary that make configurability necessary?

Management style is a big one. Across the world, you’ll find very different theories and approaches. Some are very top-down. A sales manager determines which doctors should be visited by a rep, makes visit recommendations based on priority and the rep follows them.

Then, there’s the other end of the spectrum, which is more common in countries like the United States. Reps are given more autonomy with the assumption that they’ll make whatever decision puts them closest to meeting the sales objectives set before them. Beyond issuing a prescriptive list of visit recommendations, managers empower reps with the support and information they need to do their jobs.

As you can imagine, these two management approaches are going to inspire two very different ways for reps and district managers to think through the next best thing to do.

Of course, there are also different laws, regulations and union practices that need to be taken into account. You wouldn’t want to issue a visit suggestion for a time when unionized reps are prohibited to work, for example. We have to consider all of it.

How does Aktana’s Decision Support Engine provide the framework to support this range of preferences, requirements and behavior?

Aktana’s suggestions are driven by both targets and rules. Targets are pretty straightforward. For instance, your District Manager has identified Dr. A is a high-priority doctor who should be visited twice a month. Let’s pretend it’s the 14th of the month and you still haven’t seen Dr. A. As a result, you get a reminder to do so. That’s a target-driven suggestion, which better reflects the top-down management style I first mentioned.

Aktana also makes suggestions that aren’t necessarily driven by targets; instead, they’re driven by what we’ve learned over time. These are rule-driven suggestions, which use conditional logic to determine the sequence of actions that will work best in a given situation. If a doctor suddenly starts writing 20 percent fewer prescriptions, for example, what should the rep do?

Our configurable user interface allows each pharmaceutical team to specify the rules under which suggestions should be given based not only on observations of events, but also what we see on the rep’s calendar for the future and what the rep has done in the past. Plus, it includes the data that pharmaceutical companies have purchased and put together for the reps to consider. Like, how many prescriptions does this doctor write? How has that number recently changed? What about our competitor’s drug? And so on.

That’s simplifying it, of course. Aktana goes beyond targets and rules to ensure that the right context is considered and applies machine learning to drive suggestions that make sense to the rep. Most of today’s reps are being bombarded by the amount of information they need to digest. It’s not being filtered and put into a format that’s easily accessible, synthesized and condensed. That’s the whole mission behind decision support—to look at the all of the relevant information and be able to say “For this rep, for this account, in this situation, this is the best course of action along with relevant contextualized information for the rep to use.”

During the initial configuration process, how do you ensure the DSE knows what’s relevant and what’s not?

We always start by meeting with our clients to understand their brand strategy. Most of the time, brand teams have already developed a strategy, but they don’t have the IT systems in place to automate it sufficiently—or monitor its effectiveness, in a very granular way, over time.

During those discussions, we start to determine what the rules need to be in our system in order to execute their specific strategies. Here we have to ask, “Where’s the data going to come from?” Most of the time, they have the data—and Aktana’s DSE is data-agnostic, so we’re able to consume any data regardless of channel. It’s just not always immediately in the form the engine expects, so our integration team simply reformats it to feed into the engine without losing any data integrity. Once we’ve met with the sales and marketing teams, we can configure the rules in the DSE to accurately reflect and execute that brand strategy for each market segment.

Once the configuration is complete, what’s the last step to ensure the DSE delivers suggestions reps will adopt?

A. Ultimately, the DSE is set up in a way that gives sales and marketing teams the freedom to articulate the multichannel strategy unique to each brand, while also making sure suggestions “feel right” to the reps that receive them.

We’ve built our DSE to consider contextual factors like location or physician availability to ensure the suggestions make practical sense. For example, we wouldn’t want to suggest that a rep drive three hours out of the way to visit a doctor today when he or she is scheduled to be in that area of the territory tomorrow. In addition to balancing these factors, we’ve also incorporated a learning module. So if a rep doesn’t have any visits scheduled on the calendar, we can examine the rep’s history, predict his or her pattern of travel and go from there.

In the end, it’s all about building a DSE that can be configured to reflect the real-world pharmaceutical marketing landscape as accurately as possible. Configurability allows that to happen, and our latest version of software is a testament to that. Thanks to the experience of our early deployments, we were able to anticipate our customers’ needs by strengthening rule-driven behavior to deliver a DSE that’s even more flexible, scalable and easier to deploy. The code is even structured so that future needs can easily be addressed. For example, if a team needs to add another channel or schedule out visits for the entire month, we can make it happen. I’m pretty proud of that.