All Collections
The MODELS
Data QUALITY
How to conduct a comparative analysis?
How to conduct a comparative analysis?

You want to compare data from two different providers.

Sébastien GUICHARD avatar
Written by Sébastien GUICHARD
Updated over a week ago

To conduct a comparative analysis between two data providers we need to validate several parameters.

1. The GTIN codes of the product scope.

Firstly to be able to a comparison of aggregates, we need the list of GTINs for all the products of this aggregate, indeed having access to this list of GTINs in full transparency is a right.

Openhealth makes the GTIN code available for 2 reasons:

  • For transparency and to promote interoperability. The user of the HUB platform can extract data from their applications with the GTIN Code

  • In order to know at any time the content of an aggregate (example: a product class) and to follow its evolution over time with full traceability.

Example: evolution of the content of a product class following the marketing of new products.

I draw your attention to the fact that an aggregate at OpenHealth, may have the same name at another data provider, but this does not imply that its composition is identical.

2. The data source

To be able to make the comparison between aggregates other criteria come into play, including the data source: Pharmacy, Parapharmacy, etc ...

It is not the same to look at sales data for mouthwash type products in PHARMACY and PARAPHARMACY, see CONSOLIDATED (PHARMACY + PARAPHARMACY)

3. The geographical territory

Also the geographical territory is very important, at OpenHealth we have Metropolitan France, but also Corsica, Monaco, the DROM-COM, international data (Germany, Switzerland, Italy, Portugal, etc)

And you will agree that looking at the market for solar products in December in Metropolitan France compared to data from Reunion Island, is not the same.

4. The time period

The time period is one of the parameters of the analysis, we need to compare 2 identical periods. This allows us to erase the effect of seasonality, or the presence of a one-off atypicality that can be translated by a social movement, for example the impact of the yellow waistcoats which had many economic impacts.

5. Indicators and calculation method

To complete the analysis we need to compare indicators, and to do this we need to know the associated method of calculation.

At OpenHealth, the calculation method for sales values:

This is not the same as calculating value sales from a reference price, and this method allows us to see the price disparity by product, particularly via the sheet: Price Dispersion

5. THE LAG: First sale dates and registration date

Finally OpenHealth is able to provide for each of the GTINs within an aggregate, the date of 1st sale as well as the date the GTIN was listed in our repository. This is a very differentiating point, as we are able to intercept the sales data of a product even before it is created in our repository. These sales are then recorded and then once the product is created, they are attached to its product throughout the history.

These two dates are very important in determining the lag between when we first intercept a product's sales data and when the product is created in our repository.

The difference between these two dates allows us to determine and measure the lag, and thus deduce the validity period of the data.

Did this answer your question?