UDEHA

Daniel Ek

Nearly two years of licensing talks. Then eight countries at once

When the thing standing in your way is a permission, more building is the comfortable way to avoid asking for it

Spotify

  • Brand
  • Consumer products
  • Distribution
  • Platforms

The pattern they actually represent

Spotify began in April 2006. Writing the software for a music player is a hard problem with an end to it. A catalogue is not that kind of problem. A catalogue is a stack of signatures.

Six weeks after the service opened, a reporter who had just interviewed Ek wrote the sentence this entry is built on: the service "has been in the works for nearly two years now, although it appears to have been time well spent". Spotify launched in eight European countries at once, with licensing deals from all four major labels plus The Orchard and Merlin.

Nearly two years is not how long it takes to write a music player. The blocker was never a build. It was a set of permissions held by people who could say no, and more engineering would not have moved one of them an inch.

His own account is easy to misread. He told that interviewer that instead of legislation, the way to stop file-sharing was to create a better product. He is arguing for the product — but notice what the product had to beat. Free and instant. That meant a complete catalogue, which meant the signatures.

What launched shows what he held out for: a free tier paid for by advertising, beside a day pass at £0.99 and a subscription at £9.99 a month. Free was not a discount. Spotify's filing later reported that advertising-supported listening drove more than 60% of its gross added paying subscribers.

The negotiation never ended. Ten years on, that same filing still lists dependence on third-party licences among the risks that could damage the company. The rights were not a phase on the way to a business. The rights were the business.

The blockage it speaks to

Read this one if you are stuck between two versions of your business and keep working on the product.

The scene is specific. Somewhere there is a version of what you sell that needs a partner, a licence, a distributor, or one uncomfortable conversation with a person who can refuse. There is also a backlog. The backlog is real work, it is visible, and it never rejects you. So the week goes into the backlog, and the conversation moves to next week, where it has been sitting for a while.

That is not weak discipline. Building is the part of the business you fully control, and control is what you reach for when the outcome is held by somebody else. Rebuilding also produces evidence of effort, which is exactly why it settles nothing. At the end of the quarter you have a better product and the same permission problem.

Ek is the right study for that stuckness because his version of it was impossible to hide from. Software that plays music cannot be improved into a licence. Yours is harder to see, since more building always produces something. The question underneath both takes one sentence to answer: is the thing between me and shipping something I make, or something somebody grants?

Three moves you can steal

Name the person who has to say yes. Write the outcome you want at the top of a page, and under it the name of whoever can block it — a label, a platform, a regulator, a distributor, one buyer. If the page stays empty, your blocker really is a build, and you should go build. If a name appears, put a date beside it this week. Anything that does not move that name is comfortable work.

Decide the one term you will not trade before you are in the room. Spotify's free tier survived into the launch and into every deal behind it, and it was still producing most of the paying subscribers years later. A term written down in advance turns a long negotiation into a known cost. A term you never wrote down gets traded away quietly, in exchange for finishing sooner.

Let the finished thing wait. A build that is done pressures you to ship it into whatever opening is available. Eight countries with four majors is a different company from one country with a partial catalogue, and the difference was bought with time. Waiting is only a strategy if you can say what you are waiting for and how you would know it had arrived.

Where the pattern breaks

Running this pattern hard produces a bill, and Spotify publishes the bill itself. Access is paid for out of a shared pool, and the pool is divided by share rather than priced by the play: in Spotify's own words, your streams in a market divided by the total streams there is your share of that market's royalty pool. Shares concentrate, so the curve is steep by construction. In the Loud and Clear figures for the 2025 data year, the 100,000th-ranked artist generated more than $7,300. Rank one hundred thousand is the top of the tail, not the bottom of it. A market that feels effortless on one side is being paid for on the other. Before you copy the front end of a two-sided business, read its supplier ledger.

What to do this week

Write one sentence with a name in it: nothing ships until this person says yes. If you genuinely cannot fill in the name, your blocker is a build, and this is a building week.

If you can fill it in, that name is your week. Send the first message today, before the product is ready, because the product being ready was never the condition. Ask for the specific thing rather than for a conversation — the licence, the shelf space, the referral, the release, the price.

Then write the one term you will not trade, and keep it where you will see it on the call.

Leave the product alone until Friday. It is already finished enough to be the reason you have not asked, which is the only job it has been doing for a while.

In their words

Instead of legislation, we realised that the way to stop people consuming content through P2P networks is to create a better product.
Music Ally interview · 2008
We also realised that music today is more about access to music than actually owning music.
Music Ally interview · 2008
Music was too important to me to let piracy take down the industry.
Spotify F-1 letter to shareholders · 2018
Tomorrow does not become the most important day for Spotify.
Spotify newsroom · 2018

Turning points

  1. 2006Starts Spotify with Martin Lorentzon and begins negotiating for a catalogue the company does not have
  2. 2008Opens in eight European countries at once with deals from all four major labels and the independents
  3. 2011Enters the United States as the eighth territory with ten million registered users already in Europe
  4. 2018Lists Spotify on the New York Stock Exchange by direct listing with no underwriters and no new shares

What it speaks to

Read more in the Library

Others worth studying

Sources