Skip to content

We scrapped the data from the Cdiscount website in order to do some deep analysis on price dispersions.

Notifications You must be signed in to change notification settings

Jean-Lcs/Scraping-analysis_Cdiscount_computer

Repository files navigation

Scraping-analysis_Cdiscount_computer

Cover

After having scraped all the computers available on the Cdiscount website, we started by processing the text data to isolate the different characteristics of the computer: RAM memory, storage memory, screen size, retailer etc... We then decided to focus our study on the two brands of computers most present on the site, so as not to miss any data and to obtain reliable results: Macbooks and Dells.

We were therefore able to draw some conclusions and make some assumptions. Here are some of them:

  • We can clearly see that there is a pattern emerging on price dispersion, most computer prices are either condensed around low prices or low prices but there are not many computers priced between these areas. We can interpret that people devalue the price of a good as soon as it is not new. Therefore, a small defect that shows that the computer is not new can lead to a big loss when selling.

  • We can clearly see that the price dispersion seems to decrease with the average price of the good. The graph is less smooth for macs because there is more data and therefore more randomness in some values. Wan can interpret that sellers are more careful when the price is high. This is indeed understandable, for a medium sized computer you don't make the effort to find the exact value but for a high tech computer (and therefore a computer that might have more visibility) it makes sense to do more research. Also, for cheap computers it can be complicated to find the value because of how it was used, how old it may be, etc.

  • What we call 'expensiveness' here is the average buyer's rating for different computer models. We can clearly see that the expensiveness can be very different from one model to another. Therefore, some sellers selling the same computer may set different prices, which is the reason for the price dispersion. We can interpret that some sellers have a very good reputation (either through reviews or because they are famous). Another theory is that there is a circle effect: the better their reputation, the more they can sell and therefore the more reassuring it is for the buyer. This allows them to gain reputation and raise prices accordingly.