The album lost its shelter

A compact disc player in 1986 had a button marked with two triangles and a bar. Press it and the next track started, cleanly, at the beginning. No lifting a needle, no guessing at the fast-forward. As the researcher in the study below put it, the CD made skipping "leaps and bounds ahead of vinyl or cassette tapes."¹

So the skip became free a long time ago. That is the first problem with the idea behind this piece, that the album changed when skipping stopped costing anything. It is worth staying with the idea anyway, because the way it fails points to what did change.

The thirty-second line

Today a song on Spotify counts as a stream when it is played "for at least 30 seconds."² Since late 2014 Billboard's main album chart has counted 1,500 song streams from an album as the equivalent of one album sale.³

Put those two rules together and something new appears. A listener who skips at second 29 has done more than leave. That play does not count, toward the artist's streams or toward the album's chart position. The skip is no longer a private gesture. It is a vote, tallied and priced.

A CD skip was free and invisible. Nobody at the label knew which tracks you skipped, and it did not matter, because you had already paid for the album. The album price paid for track seven whether you ever played it or not.

The data that broke the first version

The obvious prediction follows: if streaming made the opening seconds of a track count, intros should have shrunk once streaming arrived.

They did shrink, but not when the idea says. Hubert Léveillé Gauvin, then a doctoral student in music theory at Ohio State, analyzed 303 top-10 singles from 1986 to 2015. Intros that averaged more than 20 seconds in the mid-1980s had fallen to about five seconds, a 78 percent drop; tempos rose about 8 percent, and singers reached the song's title sooner.¹ He connects this to skippable services, but he also says streaming and the skip button cannot explain all of it.¹

The decline is measured from the mid-1980s, decades before Billboard counted streams.¹³ Whatever began shortening hit intros was already at work long before anyone was paid by the thirty-second play. The version of the idea in which streaming accounting started the change is contradicted.

Radio already counted

What else measures the opening seconds of a song? Radio.

A radio listener has always had a free skip: the dial. Stations never saw each listener turn it, but they did measure. By 1983 some American stations were playing short "hooks" of songs to listeners over the telephone and asking them to rate them, a practice called call-out research, to decide what to play.⁴ A hit single has lived for a long time inside a system that tests its first impression and acts on the result.

Top-10 singles are exactly the records that passed that test. So a thirty-year decline in their intros is what you would expect from a long-running counted skip, not a sudden new one.

The shelter

Here is the part the original idea was reaching for.

Radio tested singles. It did not test the rest of the album. The long instrumental opening of track seven, the interlude, the slow song placed second-to-last: none of it was ever played down a phone line to a panel of listeners. It was paid for in the price of the album, as a bundle, whether anyone liked it or not.

The album was the part of popular music sheltered from measurement. Its deep cuts could ask for patience because nobody was counting whether they got it.

Streaming removed the shelter. Every track on the album is now counted the same way, at the same thirty seconds.² Track seven has to clear the same line as the single. The hook economy that had shaped singles for decades was extended to every song on the record.

That is the proposed explanation for why the album changed. Not because the skip became free, which happened in the 1980s, and not because skips started being counted, which radio did for singles long before, but because the counting reached the songs that had been protected from it.

Longer albums, by the same rule

The same accounting explains a change that looks like the opposite of shortening. Under the Billboard rule, streams of any song on an album count toward that album's units.³ More songs mean more chances to collect them.

In 2017 Chris Brown released "Heartbreak on a Full Moon," an album of 45 songs running nearly three hours.⁵ Nothing read here shows that its length was chosen for the charts. But the incentive points that way, and an album built as a long playlist is a natural answer to a rule that counts plays rather than purchases.

Shorter intros and longer albums look contradictory. They are the same response to one rule: make every track earn its thirty seconds, and give the counter as many tracks as it will take.

The strongest objection

The strongest objection is that this gives an accounting rule too much power. Intros were shrinking for thirty years for reasons that may have nothing to do with skips: changing fashions, faster tempos everywhere, music videos, and now short-video platforms that reward a hook in the first second. The study measured hit singles only.¹ Nothing read here measures whether album tracks changed after streaming, which is what the shelter idea predicts.

That is a real gap. The shelter is a mechanism that fits the evidence, not a finding. The test would be to measure album tracks that were never singles, before and after 2014, and see whether they started to look like singles.

And plenty of artists still make long, slow, sequenced albums that do well. A rule that rewards fast openings does not forbid slow ones. It makes them cost more.

An open edge

If the album lost its shelter when the counting reached it, where does a track still get paid for whether it is played or not?

In a physical copy. A record or CD is bought whole. The buyer pays for track seven in advance, the way everyone did in 1986, and nobody counts what happens after. The old economics of the album may survive only where people still buy the object. That would make the physical album, among other things, the last place where a song can afford a long introduction.