Just over a year ago, Billboard announced a major change in the way it calculates its Billboard 200 album chart, incorporating streams and individual track sales. At the time I whined about it because I believe the main album chart should reflect which albums people are actually purchasing.

But the biggest bummer that I had not anticipated is that Keith Caulfield stopped reporting the overall album sales total in his weekly Chart Beat column. I loved that paragraph because it showed a bigger picture of the health of the album industry beyond the top ten or even the top 200 biggest sellers of the week. It was also morbidly fascinating to keep track of how low it could dip. The lowest I saw it go before the chart change was 4.05 million in July 2014.

Well it’s gone down even further since then as Caulfield revealed in an aside in his coverage of Adele’s historic sales week: “The lowest week in Nielsen history for album sales was the frame ending Sept. 17, when only 3.51 million albums were sold.”

That is not a lot of records.

Billboard’s Ed Christman has previously reported that “The highest one-week tally recorded during the Soundscan era is 45.4 million albums, in late December, 2000.” In the same article, Christman estimates that the lowest pre-Soundscan period was 1973 when weekly sales probably “totaled around 5.5 million units.”

So now the question becomes: When will weekly album sales dip below 3 million?

* * *

Another interesting fact in Caulfield’s piece is that last week’s overall album sales were 8.2 million of which 25 represented 41% (3.38 million). Numbers 2 through 100 sold 1.48 million, which leaves 3.34 million albums sold outside of the top 100. Those are all albums that sold less than 5,000 copies each. The long tail lives!