In our previous article, we discussed why healthcare organizations are moving forecasting to the cloud. There were five main challenges in spreadsheet-based forecasting that we called out – inaccuracies, inconsistencies, resource heaviness, lack of collaboration, and loss of institutional memory.
While exploring approaches that are better than spreadsheets for forecasting, a new question arises – with several platforms available in the marketplace, how do forecasters know which one to pick? Here's our take on a few considerations while evaluating forecasting platforms.
To begin with, adopting a new forecasting platform in itself is a change management exercise for the forecasting team and its leader. The platform must pose the least number of barriers to adoption and should be able to create quick wins, as John Kotter would probably advise. Specifically, every forecasting team has few peculiarities in their forecasting approach and the new platform should seamlessly adapt to them. The platform should not determine how to forecast – the forecaster should.
Next is the question of how long it takes to replicate the existing spreadsheet forecasting models onto the new platform. This has to be rapid if we are to grab the quick wins that drive a successful change management effort.
And finally, a spreadsheet-based forecaster should feel comfortable using the new platform in a span of weeks rather than months. The platform should be intuitive to a seasoned spreadsheet-based forecaster for a faster and seamless adoption.
Once we adopt a forecasting platform, we expect it to be interoperable with the forecaster's existing software ecosystem. For example, the platform should seamlessly connect to disparate data sources (epidemiology, pricing, share analytics, persistence, etc.) and interoperate, instead of the forecasters manually inserting data points. The platform should also be capable of easily sharing data with existing reporting software and export in presentations, spreadsheets, or other formats.
A new platform provides forecasting leaders with a great opportunity to ensure collaboration and governance levels of their teams are significantly improved.
The forecasting platform should provide a centralized single source of truth for the entire organization. It should create individual roles such as team leader, reviewers, forecasters, senior leadership, reporting teams, etc. and provide appropriate role based access to them. Finally, it should enable workflows to manage the forecasting process.
A forecaster would not adopt a forecasting platform which is a black box – one that takes inputs and delivers outputs without knowing the data conversions and forecast calculations happening within the system. The platform must instill trust among the forecasters that it is executing the right algorithms on the right data to generate the right outputs. A good forecasting platform should foster transparency with a holistic view of data and algorithms that build the forecasts.
Lastly, how feature-rich is the platform? Ideally, a forecasting platform's features should go beyond mere replicating the spreadsheet functionality on software. Here are some possible features that make a forecaster's job easy – a graphical interface to build forecast flows, linking up forecasts that share common epidemiology data points, portfolio roll-up and aggregated reporting of forecasts across indications and geographies, comparison of forecasts, multi-product forecasts, and easy modeling of events.
A carefully selected forecast platform with the above considerations can not only make forecasting more accurate and consistent, less time and effort consuming, and more collaborative, but also free up the forecaster for the all-important task of business-critical insight generation.
Many #FutureReadyHealthcare organizations are switching to cloud-based forecasting platforms that best align with their objectives to generate quick, easy, and accurate forecasts. What parameters would you consider for a good forecasting platform? Please share them here.