By Toni Cejas
Aktana presents a three-part series focused on dispelling common AI misconceptions and how to engineer success that scales. In Part 3, we analyze past implementations to outline the key characteristics of every successful project. Catch up on Part 1 and Part 2 of our series here.
Artificial Intelligence (AI) is becoming one of the main pillars in the growth strategies of multiple industries, including life sciences. The technology is helping numerous organizations transform their business models, driving them to become more effective, and consequently, more profitable.
If you’re considering implementing AI globally within your organization, this post is for you. With ten years of experience and more than 300 deployments in the rearview mirror, we’ve learned what works—and what doesn’t. Here are the common traits shared by our most successful global AI projects, distilled into four essential tips.
1. Implement a Consistent Data Strategy
It is common knowledge that the main benefits of AI come from a solid data foundation: better data in, better insights out. Preparing the data that will feed your learning algorithms is vitally important. But in our experience, it’s also the step in the implementation process that eats up the most time and has the most room for improvement.
If the goal is to scale quickly and efficiently, any data sources that will be used should be as consistent as possible across the organization. Let’s look at one of the main data sources in the pharma commercial operations environment—the CRM—as an example:
It’s common practice to use the same CRM platform across the global organization. But due to the inherent flexibility of such platforms, it’s not uncommon to find different uses of your data model in different countries, or even from one sales team to another. This is not an impediment to the implementation of AI-based strategies, but it does slow them down. Creating standards in CRM implementation accelerates the development of AI algorithms and their subsequent deployment in multiple geographies.
The same principle applies to the other data sources that AI models will leverage. Having a mature data infrastructure in place to deliver data in the right formats, in an industrialized and homogeneous way across geographies and markets, will save countless hours on the front end of any implementation. This implies the existence of Master Data Management and Data Warehousing processes and technologies that include, if possible, most of the data sources to be used.
2. Start with the Essential, Move to the Sophisticated
Our most successful projects have never taken a “big-bang” approach. Instead, they’ve grown in small steps, starting with fundamental use cases that have broad-based appeal and help end-users adjust to new working methods. As users grow more comfortable over time, they move on to more sophisticated use cases that address the specific needs of much smaller user groups.
To achieve this goal, we recommend designing use case libraries—a series of templatized customer micro-journeys (e.g. event follow-up or consent gathering)—and data management workflows in layers. The first of these layers should contain those use cases common to the entire organization, corresponding to strategic business practices for the company. The next layer would focus on region-specific practices, perhaps stemming from different legal frameworks, the existence of different data sources, etc. The next layer would focus on the particularities of countries within the region or sales teams within the country.
Creating a hierarchy of configuration layers allows for rapid deployment of new projects, making it possible to get complete implementations up and running in a short time and with little effort.
3. Establish Global Governance for Implementations
In global projects, it is essential to have repetitive, almost industrialized implementation processes that allow us to optimally coordinate Aktana’s teams with their client-side counterparts. All of our most successful projects have started with the joint design of these implementation processes and their subsequent dissemination to the different regional teams involved in the project.
Global Governance teams are an essential vehicle for mobilizing the necessary resources in each region and country, conveying the strategic importance of the program, and providing the business case justification and intended business impact. And perhaps most critically, they help to establish visibility on regional and market requirements that serve to stimulate local engagement and to ensure that the broader platform is tuned to deliver effectively on the ground.
4. Make a Total Company Commitment to Embrace Your New Initiative
Change generates restlessness, and restlessness generates resistance.
Transformational technologies and strategies like AI / ML or Next Best Action can have this effect on a large number of end-users, whose involvement and support are key to the success of the project.
As you embark on this type of global initiative, a focused, forceful message from the company’s top stakeholders can go a long way in driving adoption. Clearly express the long-term commitment to the project and the different benefits that this change will bring. Discomfort is temporary, but the gains achieved through digital transformation can set up your organization to succeed for years to come. Invest in change management through training and dedicate internal communication to ease teams into a new way of working that will ultimately help them deliver a superior customer experience and bring new efficiencies to their work.
When in Doubt, Plan for the Possibility of Going Global
More than half of companies expect to see AI as a piece of digital engagement nearly ubiquitous throughout healthcare, including the pharmaceutical industry, by 2025. Do not be left behind, waiting until all of the perceived requirements are in place to get started. Establish careful governance over the AI initiative and then make a phased global roll-out part of the plan. The platform will learn and evolve in concert with the business. After all, that’s exactly what AI is designed to do.