The Commensurate Sales to Popularity Concept (CSPC) : Introduction

04 Scale and Balance Eric Johnson


Hello everyone! Let me introduce this third Master Series with this few words: Facts can also lie.

Katy Perry sold more records than Dire Straits, Bon Jovi and the Guns N’ Roses. White Christmas by Bing Crosby and Candle In The Wind by Elton John are the two bestselling singles of all-time. Mariah Carey album Butterfly sold more copies than any album ever released by Taylor Swift.

Yes, something is wrong here. All those claims are correct in absolute terms but completely meaningless. The natural thinking when watching a bestselling records and artists list is to assume it reflects the most successful ones. This is false. As fitting as it seems, just checking raw data isn’t enough to identify which songs, albums and artists are popular. Most articles state “record sales” mixing everything, as if one digital track sales and one deluxe 4CD album was the same product. Lately, streaming has been growing too. Many voices denied them as legit ‘sales’. In reality, they highlighted a need that has been around for decades but ignored for too long. Indeed, music industry needs one single relevant and consistent scale that would fit for all records, from all eras, rather than barely looking at ‘units’ although they refer to apples and oranges.

You got the point? Wait, there is more. If format differences are important, we also need to consider the way majors have been working out on a catalog. If you have a look into Elvis Presley catalog, you will easily notice it has way more compilations than studio albums. Thus, checking studio albums sales give the feeling he was a pretty weak seller. The point is that his singles were available over a large amount of records, thus diluting sales over all of them. In the other side, various one-album wonders a la Ace Of Base look like monster sellers, barely because all their sales are concentrated on a single product. As a consequence, our generic scale must also put on par all recordings independently from the way they have been exploited.

This Master Series will be presenting figures for various albums using a completely new point of view, aimed to be the figures reflecting the best the real popularity of albums and artists.

This generic scale is named the Commensurate Sales to Popularity Concept, abbreviated CSPC. When displaying figures using this concept, I’ll be using the measurement unit “CSP”.

9 thoughts on “The Commensurate Sales to Popularity Concept (CSPC) : Introduction”

  1. Wonderful and innovative approach. Only possible thanks to tools like Spotify!

    I’m very old fashioned, so, for me, it will be hard at first, hehe.

    Shouln’t music videos count as a slightly more “important” sale than one album’s? I mean, I think they are slightly more expensive and have more content. (Michael Jackson fans will be glad, The Beatles’ ones will complain!).

    1. Hello Hernán, thanks for the nice comments!

      Indeed Spotify is an incredibly powerful tool giving a very good understanding of what’s really popular. We are lucky to have it available!

      As for Music Videos, they are indeed priced often a bit higher than an album. I have put them on par mostly because they are almost always released with one live CD with people buying one or the other, rarely both. Thus, it is a 1 to 1 relationship. Obviously compilation of music videos (rather than a live concert video) are different but in any case they will be treated the same way as compilation albums.

  2. Interesting ideas and I look forward to seeing how things are impacted as a result.

    One point where I differ (but this is only a matter of personal opinion) is on video sales. I just think it’s a completely different medium. In practical terms few music fans would buy a video instead of a song/album. They would buy it as well as the song/album, either because they are fans and want to collect items, or because the video adds some extra dimension that is enjoyable over and above the song (in which case is it really the song itself which is more popular when the video is sold). Just as you wouldn’t add T-shirt sales to record sales, then I would say the same for videos.

    But as I say just a personal opinion and I realise there is a case for inclusion.

    1. Hello Davyboyb, nice to see you around!

      First results will come this week itself, I hope you will enjoy them! About music videos, the fundamental difference with a T-Shirt or a Poster is that it still contains the music. A second important point is the place they own on a catalog. Music videos are hardly stand alone releases, they are packaged along with a Live album for the large majority of the cases. Thus, it ends up being only an alternative to the album, Just like a Youtube view is a valid alternative to a Spotify stream.

      Obviously, this will bring in some extra sales due to fans buying both the CD and the DVD of such releases, but that is also a trademark of the popularity of the material if it is big enough to create fans. They are on par with sales of anniversary / expanded editions of classic albums.

      As you said it is a matter of opinion yet and does not bring drastic changes to results so everyone can make up his mind about it.

  3. Thanks MJD. Like you say the videos shouldn’t make dramatic changes in most cases, so I guess I’ll just agree to differ.

    Look forward to results.

  4. Hi MJD,

    Great job but I have the feeling you don’t take into account youtube videos in you calculation of comprehensive streaming : you apply the same coefficient (roughly, spotify streams = 38% of comprehensive streaming) to non-single tracks and to single tracks wether they have a music video or not.

    Nirvana’s Smells Like teen spirit has already 612M streams with spotify and the vevo music videos combined only, more than you give it to the song in your post about Nirvana.

    As you said,”a Youtube view is a valid alternative to a Spotify stream.” so shouldn’t the official music videos – at least for singles – be taken into account ?

    Another question : what’s the market shares of spotify in France ? Can we apply the same 0,38 coefficient to have the french comprehensive streaming (mainly for french acts) ?

    Thanks a lot

    1. Hello zrthur!

      You are right, only music streams are counted for into all figures of CSPC articles. I’m hoping I’ll add video streams into the calculation soon, but for that I have to wait for the value gap highlighted on 2016 IFPI Year End Report to be fixed. In concrete words, while streams on the likes Spotify bring money to the music industry, streams on Youtube bring money to… Youtube. As of now they are hardly contributing in the music which is why I’m excluding them. Several key figures in European parliament are trying to create a legislation to solve this problem, hopefully it will happen soon and our favorite artists will then start benefiting for their plays on the famous video website.

      About Spotify in France, as of last SNEP report it represents only 11% of the market in France in terms of revenue. Considering the unfavorable free to paid subscribers of Spotify, the Swedish giant share likely climbs to 15-25% if we consider the amount of plays rather than revenues. Most French artists achieving notable Spotify plays are those managing some appeal in other markets like Stromae or Maitre Gims. This latter for example has 28m streams with J’me Tire, 14m with Sappés Comme Jamais or 12m with Bella, with Est-ce que tu m’aimes? which is definitely not as big as the previous three in France is up to a massive 40m thanks to its Italian and Danish success.

  5. Hello MJD,

    The CSPC method is a great tool to measure the popularity of an album. So I think about a method to estimate the success of a song. I have in mind several formulas, and I would like your opinion about that:

    1. Album / compilation / live / video… : SongSales = AlbumSales * SongStreams / AlbumStreams
    2. Physical single: SongSales = 3 * SingleSales / NumberOfSongs
    3. Digital single: SongSales = DigitalSales / NumberOfSongs
    4. Streaming = SongSales = SongStreams / 1500
    With NumberOfSongs = number of songs from the studio album where was released the song (or 10 if Orphan).
    So, the total “sales” of a song (its success) will be the sum of the four parts mentioned above.

    Do you think this methodology is consistent? If not, what would you replace or add in it?

    The easiest way to do this work is to use the figures you post in your CSPC analysis.
    I have already done this job for Adele (an easy case), and I could send you the Excel file (an example is always clearer than explanations!).
    If you agree with that method, I could post the figures I get in each one of the CSPC articles.

    Thanks a lot to read that and answer me 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *