The advantages of total market forecasting

Robert Carbone, CEO of Futurion, and Vyanti Joseph, decision science analyst at Solvay Pharmaceuticals US, explain how total market forecasting can help predict the evolution of the market itself.



Robert Carbone, CEO of Futurion, and Vyanti Joseph, decision science analyst at Solvay Pharmaceuticals US, explain how total market forecasting can help predict the evolution of the market itself.



Traditional product forecasting looks at a relatively undefined market, says Robert Carbone, CEO at Futurion.


Established models are essentially based on a supply chain approach, including ex-factory sales data, wholesaler data, and the various distribution channels.


The customers for such data are principally the supply chain itself and finance.


In contrast, market forecasting needs a total approach, according to Carbone, for which the main additional factor is competitors.


Sales data sources will vary according to geography, depending on the healthcare system.


For example, in the US, TRx audit data will commonly be used while in Europe IMS standardized unit data will commonly be used.


Total market forecasting


Carbone explains that total market forecasting (TMF) encompasses predicting not only your own likely sales but also those of your competitors as well as the evolution of the market itself.


The process is driven by assumptions about what might happen.


For example, market assumptions concern trends in the overall market while share assumptions concern how the market is divided among competitors.


In practice, the forecasts will be broken down to a detailed level, even for each of the packs for a product.


Carbone identifies several types of forecasting activity in a typical company.


Volumes need to be planned in the short term, while strategic planning needs long-term estimates.


Underlying these activities will be the assumptions and growth drivers as well as budget considerations, which need frequent revision, leading to the current preference for rolling forecasts.


Requirements for active product ingredients must be included in estimates, Carbone says: It's impossible to harmonize all these activities without a total market forecasting approach.


Long-range focus


At Solvay Pharmaceuticals US, an independent group carries out forecasting, explains Vyanti Joseph, the company's decision science analyst.


A major annual forecast, which has a long-range focus and sets the budget for the following year, is created every third quarter.


The overall budget, of course, determines the various departmental budgets that control the whole sales and marketing function, which is broken down to monthly and unit levels.


The monthly figures trigger examination of trends in prescription patterns and other factors.


Every quarter, production forecasts are revised.


The whole process is readily managed using a spreadsheet system, explains Joseph, with a top-down approach primarily driven by market share data: The Excel model has been within 3% of accuracy.


Yet Joseph is conscious of the limitations of a spreadsheet model, good as it is.


The need for transparency


A key problem is transparency to the customers using the data, she explains; if customers are unable to see how the output is derived, they are less confident in it.


An Excel system can be very large and complex, difficult to use, and can have major problems of user accessibility and version control.


Interfacing with other systems needs a lot of manual work.


These and a host of other issues led to Solvay Pharmaceuticals USs association with Futurion.


Migrating to a single TMF system was not without its challenges, according to Joseph.


There was a lack of internal consistency at Solvay among databases, which necessitated extensive customization of the system.


Nevertheless, Joseph says, We are getting there, with much of the interfacing automated. It's nice to have all the history in one place.


Fast, effective access


Carbone says using an Oracle database for the system enables much faster and more effective access to reports on market share.


The user can input multiple assumptions to generate what if scenarios and manipulate reports on the fly.


For example, different supply channels such as long-term care can be selected to see the effect of a new competitor on a particular part of the market.


Also, the scenario of interest can be expressed in a wide range of forms, such as unit sales (packs), dollar amounts, and prescriptions.


There is no need to interrogate multiple sources, Carbone explains, and by accessing different views of the data the forecasts can be corrected and verified.


Perhaps most importantly, centralizing the process in this manner makes it accessible to non-specialists.


This article was adapted from a talk given at Forecasting USA 2009. For information on Pharma Forecasting Excellence Europe, click here.