TBD: Preparing for the future of Pharma: a data informed framework for strategic workforce planning
Poster Presenter
Raul Mora
Senior Strategic Operational Intelligence Head
Hoffmann-La Roche Switzerland
Objectives
Many external factors are reshaping the future of Pharma and healthcare, the next 5 years will redefine how the industry’s workforce will look like. The SWP framework aims at defining a data-informed approach that provides recommendations to build the best strategy to be future-ready
Method
The framework was developed for the Pharma division of Product development across multiple functions, during 2020-2021.
Multiple sources of data were collated and curated to establish the baseline for Body of Work and Skill Clusters. Through a mix of data analysis and SME interviews
Results
Multiple portfolios of work and projects were curated and aggregated to have a harmonized view of work done across the organization, aiming at having equivalent pieces of 'skill demand' across multiple domains, yet transferrable.
Through mining of ERP systems and local trackers cross checked with SME interviews, a baseline for talent capacity and proficiency was developed. Where gaps were present (data not available or quality not good enough), an approach of synthetic data created through distributions of smaller subsets was created. This allowed for the framework to have an early baseline that can be refreshed, updated and revalidated at any given point where robust data becomes available.
The framework then developed a recommendation engine for workforce strategies that would aim at giving an enterprise wide consistency of how to approach the current talent gaps in the organization. It also aims at establishing an early approach to shift towards skill based talent flow as opposed to role based. The engine was designed initially with a deterministic set of rules for recommendations, but subsequent iterations will contain stochastic elements that will allow to account for external, unforeseen shifts in workforce market conditions.
The recommendations were then presented along with a tradeoffs matrix (cost, quality, speed of gap closure, compliance) that was then presented to decision makers who would then confirm that the set of recommendations was reasonable according to their expert judgement.
Pilots to monitor the accuracy/success of such recommended (and selected) strategies are in planning, and are expected to be ready for analysis before the end of 2022.
An early concept to include Machine Learning algorithms that would allow models to be trained as the strategies are executed an validated, is expected to be ready for testing in early 2023. The data from actual strategy execution, would be used to test and retrain the models.
Conclusion
Although most workforce decision makers seem to rely heavily on expert advise (intuition) for their local strategic workforce gaps, it was widely acknowledged that this intuition worked less accurately with areas where they were not experts.
In addition to this, when their local org or skill domain is contrasted/complemented with other org areas, data informed insights become potentially more significant than just local decisions. Overall, enterprise-wide strategies require of a structured framework that allows the business to make truly enterprise-wide strategic workforce planning.
Shifting from a Job Role based workforce strategy, to a pool of skill based workforce increases the complexity of establishing work demand and workforce capacity. This is especially true, when new value propositions are being developed, and the future of work/workforce will likely imply a need for individual development journeys.
With pharma/healthcare work becoming increasingly complex (do more with less, adoption of personalized healthcare, value-based assessments, increased regulatory requirements), it is paramount to have a robust, yet flexible approach to strategic workforce planning. It is unlikely that this will be achieved without data-informed models that can keep track of higher volumes of data, more granular level of analysis and remain flexible in an ever more complex setup. Shifting early enough to a skill based approach enabled by a Machine Learning modelling will allow organizations to consume, process, analyze and collect big amounts of data and generate the right level of insights for leaders to make informed decisions.
