What are the different steps to calculate sector sales at OpenHealth?
Sales to the sector are the result of a data processing process that can be broken down into 3 stages:
- Step 1: OpenHealth calculates national sales data, through an extrapolation process, the modalities of which we detail in this article
- Step 2: Sector reallocation coefficients are then calculated for each product and sector
- Step 3: National sales data are reallocated in each of the sectors thanks to the sector reallocation coefficients
So we have National - > sector reallocation coefficient > Sector
Modeling national sales and then reallocating them to the sector is a more robust method from a statistical point of view than extrapolating to the sector and reconstructing national sales. (sector - > national)
Indeed, the overall statistical margin of error of national extrapolation is much lower (due to the greater number of pharmacies) than the sum of the margins of error of the extrapolations sectors.
Why do production operations (MAM, IPR) affect sales?
The data model at OpenHealth is designed to dynamically reflect the commercial reality in pharmacy: product launch (IPR), product deletion (SPR), categorical approach (MAM) . The hierarchical criteria of the product scope of your market (segment / brand / product) are therefore dimensions taken into account in the calculation of the sectoral reallocation of your sales.
Let's take the example of an IPR, which is the addition of a product in a market.
This addition will have impacts on all products on the market, even on those that are not affected by the IPR. For example, adding product A to market X will change the market share of all products in the market.
In the same way, the introduction of a new product will modify the relative weight of the sectors, which is a variable taken into account in the coefficient of reallocation of sales from national to sectoral.
historical data is recalculated at each production operation, an RPI and MAM can marginally influence the allocation of national sales to the sector over all periods and across all products.
Why may this effect not be visible in other panelists?
OpenHealth has the particularity of processing all the data collected in flow, unlike historical models for sending Adhoc files on Excel which freeze the data at an instant t.
This difference explains why data can change over past time periods ('backdata') when market definition changes.
Working on data in flow actually has many advantages, such as being much more reactive in updating data (D + 1) and therefore being in touch with market events.