Farnam Street helps you make better decisions, innovate, and avoid stupidity.

With over 400,000 monthly readers and more than 93,000 subscribers to our popular weekly digest, we've become an online intellectual hub.

John Seabrook and the Modern Song Machine

Ever noticed a certain “sameness” to the pop songs you hear nowadays? A similarity in their structure, feel, and the voices you hear on the tunes? You’re correctly clued in.

Mostly gone are the days of Elton John and Bernie Taupin sitting down at a piano to work out a “Bennie and the Jets” — crafting the chord progression, the melody, the rhythm, the arrangement, and finally, the lyrics.

The pop music running the airwaves today, those songs you seem to know but aren’t quite sure how you know them, are created in an interesting and deliberate way. Depending on your stance, the result is either horrifying or fascinating.

The method is called track-and-hook songwriting, and employs an entire industry of sub-specialists whose job is to put together pieces of a Frankenstein beast that ends up as a monster hit. Like many other industries, songwriting has been changed immensely by the Internet: With attention spans shorter than ever and avenues for music consumption unlimited and mostly free, hits are more important than ever, not less.


This interesting result wasn’t well predicted. Chris Anderson, of The Long Tail fame, predicted that the Internet would lead to less hit-domination and more exploration of individual passions, writing in 2005: “If the twentieth century entertainment industry was about hits, the twenty-first will be about niches…This is not a fantasy. It is the emerging state of music today.”

He was wrong, though, as John Seabrook writes in his marvelous book The Song Machine: Inside the Hit Factory:

Nine years after The Long Tail, the hits are bigger than ever. Of the 13 million songs available for purchase in 2008, 52,000 made up 80% of the industry’s revenue. Ten million of those tracks failed to sell a single copy. Today, 77 percent of the profits in the music business are accumulated by 1 percent of the artists. Even Eric Schmidt, the CEO of Google and an early supporter of long-tail theory, changed his mind. “Although the tail is very interesting, and we enable it, the vast majority of the new revenue remains in the head,” he said in a 2008 interview with McKinsey, the management consulting firm. “In fact, it’s probable that the Internet will lead to larger blockbusters, more concentration of brands.”

In order to fulfill the demands of a hit-driven business, threatened and changed by the democratization of music consumption, the business has evolved to create hits the way McDonald’s creates hamburgers: on an assembly line.

Pon de Replay

Pop stars don’t generally write any of the music they become famous for. They are faces, personalities, and voices, though with the advent of digital music-making and the ability to digitally auto-correct vocal performances, the latter is increasingly unimportant.

The whole thing starts by crafting the track, which is done by the producer and the topliner. John Seabrook lays it out in his book:

By the mid-2000s, the track-and-hook approach to songwriting–in which a track maker/producer, who is responsible for the beats, the chord progression, and the instrumentation, collaborates with a hook-writer/topliner, who writes the melodies–had become the standard method by which popular songs are written. The method was invented by reggae producers in Jamaica, who made one “riddim” (rhythm) track and invited ten or more aspiring singers to record a song over it. From Jamaica the technique spread to New York and was employed in early hip-hop. The Swedes at Cheiron industrialized it. [Ed: Songs like “I Saw a Sign” from the early ’90s.] Today, track-and-hook has become the pillar and post of popular song.

Why do it this way, instead of the old way where one or two people wrote a song and the singer put personal lyrics on top of it? Seabrook thinks it’s because parts of the songwriting process have become extremely specialized, a sort of natural selection process has occurred where songs must be engineered for maximum addictiveness to survive. As they figure out a successful formula, it’s applied over and over to create more hits:

As a working method, track-and-hook tends to make songs sound the same. Dance music producers have always borrowed liberally from others’ grooves. There’s no reason not to: beats and chord progressions can’t be protected under existing copyright laws, which recognize only melody and lyrics. As dance beats have become the backing tracks to a growing number of pop songs, similar-sound records have proliferated. The melodies themselves are supposed to be unique, but because of the way producers work with multiple topliners, tracks and melodies tend to blur together.

In 2009 for example, both Beyonce and Kelly Clarkson had hits from tracks written by the superproducer Ryan Tedder. One was Beyonce’s “Halo,” which peaked at number five in May, and the other was Clarkson’s “Already Gone,” which got as high as number thirteen in August. Clarkson wrote her own top line, while Beyonce shared a credit with Evan Bogart. When Clarkson heard “Halo,” she thought it sounded too much like “Already Gone,” and feared the public would think she had copied Beyonce’s song…But nobody cared, or perhaps even noticed; both songs were hits.

The songs are engineered precisely to hook the listener as soon as possible and then re-expose them to a hook over and over. Attention spans are too short to allow “dead space” in a song:

In a track-and-hook song, the hook comes as soon as possible. Then the song “vamps”–progresses in three-or-four chord patterns with little or no variation. Because it is repetitive, the vamp requires more hooks: intro, verse, pre-chorus, chorus, and outro hooks. “It’s not enough to have one hook anymore,” Jay Brown explains. “You’ve got to have a hook in the intro, a hook in the pre, a hook in the chorus, and a hook in the bridge, too.” The reason, he went on, is that people on average give a song seven seconds on the radio before they change the channel, and you got to hook them.

Once there is a hooky, chorusy, bubbly tune to draw the listener in and keep them there, the singing comes in next. In today’s pop world, vocal quality no longer carries the importance it once did, nor does writing meaningful lyrical content. The lyrics must have a meter that fits the rhythm of the song, and the vocal itself will be heavily processed and engineered by the time the song hits the airwaves. Dozens of takes will be digitally cut-and-pasted together to create the final song. This process explains why glossy pop seems to carry a similar sound and feel even if the songs are recorded by artists with totally different singing voices.

As an example, Seabrook describes songwriter Ester Dean laying down a vocal track for a potential hit:

Dean was dimly visible through the soundproofed glass window, bathed in greenish light. She took out her phone, and as the track began to play she surfed through lists of phrases she copied from magazines and television programs: “life in the fast line,” “crying shame,” “high and mighty,” “mirrors don’t lie,” “don’t let them see you cry.” Some phrases are categorized under headings like “Sex and the City,” “Interjections,” and “British Slang.”


Grabbing random words out of her phone also seems to set Dean’s melodic gift free; a well-turned phrase would restrain it. There was no verse or chorus in the singing, just different melodic and rhythmic parts. Her voice as we heard it in the control room had been Auto-Tuned, so that Dean could focus on making her vocal as expressive as possible and not worry about hitting all the notes.


“See, I just go in there and scream and they fix it…” she tells me, emerging from the booth, looking elated, almost glowing. She touches the back of her arm, feeling that million dollar chill.

Moneyball-style, the engineering of musical addictiveness takes an old pop-music concept — giving listeners heavy exposure to a song so it becomes familiar — and uses that to predict which songs will be hits:

The main difficulty Zapoleon had to overcome in creating Hit Predictor [Ed: a music testing service], he says, was that people don’t know if they like a song unless they’ve already heard it. “There’s an old adage that you can only do research on people who are already familiar with the song,” he says. Zapoleon refers to this as the “rule of three” –you have to hear a song three times before you know if you like it or not.


Zapoleon’s solution was to replicate the rule of three in a two-minute remix of the song. “We take the thirty-second meat of the song,” he explains, “which is generally the chorus but sometime it’s not. And then comes a one-minute version that has the hook in it. And then we come back again to the thirty-second hook, what I call ‘the filet mignon.'” Zapoleon’s online respondents hear the essence of the song three times, all in the course of two minutes.

In the course of a few minutes, these music-testing services can predict how likely a song is to become a hit; computer algorithms can use that data to analyze whole albums of tracks mathematically to see whether their particular combination of hooks and beats can become popular with the right marketing “push.”

A desirable track, built piece by piece inside the Machine, will be fought over by a score of pop-music icons, with a single song having the ability to launch, build, or repair multi-million dollar careers. And so the Song Machine cranks on.

Still Interested? Read the whole book for the story of modern pop.