Chartwatch Index

Corrections to my article in CW80:
Weekly Record Sales 1994--2003

After thorough inspection of my article in CW80 (p23), I have to make a few corrections and add some comments:

Due to a formatting error that I made when gathering all the data from various files, the columns are wrongly headed for the years 1998 to 2001. The second column contains the figures for total albums, and the fourth column is the sum of total albums plus compilations—rather than artist albums itself. For some reason I made that error only for the years 1998–2001, the 2002 and 2003 figures are all correct.
For instance for the first week of 1998 the correct entry with its headings should have read therefore:

WeekDateSinglesArtist AlbumsCompilationsTotal Albums
1. 10.01.1998 743,741 2,311,420 357,707 2,669,127

The singles figures for the last two weeks of 1996 are also incorrect. The correct entries should read:

WeekDateSinglesArtist AlbumsCompilationsTotal Albums
51. 28.12.1996 2,154,400 8,828,700
52. 04.01.1997 1,292,600 3,531,500

I would like to apologise to all Chartwatch readers for these arrors. I have also to stress that these weekly estimates may diverge from the actual sales figures as measured by OCC. Until a few years ago, official weekly sales figures have been reported only very rarely, if at all.

When I began this excercise in 1994 I started with the very sparse data available then, initially supported by the (much higher) trade delivery data from the BPI. The purpose of all this was to provide a pattern of sales throughout the individual year. Recently OCC has published annual totals for some years: in comparison to these, some of my estimate-based annual totals are more than 10% lower for 1994--1996. However by the year 2000 my numbers are within 1 per-cent and often within 0.2 per-cent.

There is a specific problem with one year: For 1996 my entire calculation for singles sales is based on one week's official figure. In other years there is multiple data to corroborate singles sales, but that year there was none at all. My total for 1996 is 17% below the newly published OCC number; compare that to the the divergeances for 1995 and 1996: 0.9 and 0.4 per-cent, respectively.

In light of the new figures published by OCC I plan to revise my estimates for the early years in order to bring the two sets of data in line with each other.

Back to the Introduction

Back to the Zobbel Homepage

[Created 30.05.2005, last revised 30.05.2005]