Founder stories
Sivers writes that he decided to sell on January 18, 2008 and that the paperwork took another seven months, so the closing date here is derived from his own account rather than a date he states. He had moved CD Baby and HostBaby into a charitable trust before the sale, so Disc Makers bought from the trust and the $22 million went to music education.
Online store that sold independent musicians' CDs direct to fans, paid artists every week, and charged a flat setup fee plus $4 per sale.
How Derek acquired customers
Tools used to build CD Baby
A friend asked Derek Sivers to sell his CD too. He said sure, charged $35 to add an album and $4 per sale, and CD Baby was profitable in its second month.
Derek Sivers was 27 and making a full time living from music when this started. He played gigs across the United States and Europe, produced records, ran a small studio in Woodstock, New York, and worked as the musician and MC for a circus. He had pressed a CD of his own songs and sold 1,500 copies at his shows. What he could not do was sell it online. In 1997 no business would put an independent musician's CD in an online store. The big online record stores told him the only route in was through a major distributor, and distribution was a closed system: hard to get into, slow to pay, and quick to drop you if the first few months were soft.
So he built his own checkout. In 1997 that meant $1000 in setup fees for a credit card merchant account, three months of paperwork, and a bank inspector visiting to confirm he was a real business. He had never programmed, so he copied examples out of a book until a buy now button worked on his band's site.
Then a musician friend asked if Sivers could sell his CD too. It took a couple of hours to add him. Two more friends asked. Then strangers started calling, saying their friend Dave had told them Sivers could sell their CD. He said yes to all of them. Two people who ran music mailing lists, Bryan Baker of Gajoob and David Hooper, told their readers, and fifty more musicians signed up. Sivers writes that he had accidentally started a business, and that he did not want one.
His response was to write down what a distributor would look like in a perfect world, from a musician's point of view: pay me every week, show me the name and address of everyone who bought my CD, never kick me out for selling too little, never allow paid placement. Those four points became the mission. He named it CD Baby.
Pricing took one conversation. He walked into the record store in Woodstock and asked how their local musician shelf worked. The answer was that you set the price, they keep a flat $4, and they pay you every week. He went home and wrote the same words on cdbaby.com. Because adding an album took him about 45 minutes, he charged $25 for it, then raised it to $35 a few days later because $35 felt about the same to the customer and left room for discounts. Those two numbers, a $35 setup fee and a $4 cut per sale, stayed the whole business model.
A note on the dates behind the milestones here. Sivers describes the CD Baby decade as 1998 to 2008, writes that he first built CD Baby in 1998, and says that in 2001 CD Baby was three years old. The milestones count from that 1998 launch of cdbaby.com. He says his second customer, a buyer in the Netherlands, bought on the day he launched the site, so the first sale landed on day one. On money he is equally specific: he spent $500 to start, earned $300 back in the first month, made $700 in the second month, and was profitable every month after that.
The first version did almost nothing. It listed a few CDs, each with a buy now button. Clicking it put the CD in a cart, asked for your details, and emailed them to Sivers. That was the entire site for the first year, and it was enough to be profitable. He built it in a few days because the checkout for his own record already worked.
The plan changed almost immediately, and from the outside. That second customer emailed a week later to ask about new releases. Sivers did not understand the question, because he thought he was running a credit card service, not a shop. The reply was that the customer had assumed it was a store. So Sivers rebuilt it as one.
He never raised money. He turned down the weekly calls from investment firms through the dot com boom, telling them he wanted the business smaller, not bigger. When he could not afford a programmer he bought a $25 book on PHP and MySQL and learned it. The desks were planks on cinder blocks and he assembled the office computers from parts.
Two things nearly went wrong. In 2001, three years in and with eight employees, he was working 7am to 10pm seven days a week and every decision still went through him. He fixed it by answering each employee question in front of everybody, explaining the reasoning behind the answer, and having someone write it into a manual. After two months the questions stopped, and he moved to California to make the point permanent. The second problem was expensive. Years earlier he had signed his father's paperwork without reading it, and a bank teller had talked him into running CD Baby as an alias on his existing company account rather than opening a new one. Four years in, his accountant told him his father's company owned 90 percent of CD Baby. Buying it back cost $3.3 million.
By the end, his own site describes CD Baby as the largest online seller of independent music, with over 150,000 musicians and $100 million in sales. In May 2003 Apple invited him to Cupertino to put the catalog into iTunes, 5,000 musicians paid $40 each in advance for the delivery service, and that $200,000 paid for the equipment and people needed to do it.
He decided to sell on a single day, January 18, 2008, after a few hours of writing in his diary and one long conversation with a friend, and the paperwork took another seven months. Before the sale he moved CD Baby and HostBaby into a charitable trust, so Disc Makers bought the company from the trust rather than from him. That turned it into $22 million for music education.
The business was made of something he already owned. He had paid $1000 and waited three months for a credit card merchant account, and CD Baby started as letting other musicians use it for a fee.
He set his prices by asking the local record store how their shelf worked and copying the answer. A $35 setup fee and a flat $4 per sale stayed the only two numbers in the model for ten years.
The first site put a CD in a cart and emailed him the order. That was all it did for a year, and it was profitable from the second month on $500 of spend.
The product changed because a customer misread it. A buyer in the Netherlands asked about new releases, said he had assumed it was a store, and Sivers rebuilt it as one.
He wrote his terms from the musician's side of the table, published them, and repeated them everywhere: weekly payment, buyer names and addresses shared, nobody dropped for low sales, no paid placement.
Refusing funding kept every decision his. He told investment firms he wanted the business smaller, and bought a $25 programming book instead of hiring a developer.
See what it covers before you sign up: the pattern worth copying, the channel to test, and the risk to avoid when you adapt this story.
$100 MRR
Word of Mouth
Action checklist
See what it covers first. Create an account only if you choose to unlock.
Inspired by Derek's journey? Generate a business idea in the E-commerce space using AI and real founder data.
Derek achieved 2 milestones on the path to $100 MRR
$300
The journey, decisions, and context behind this milestone
See the complete breakdown: launch strategy, validation methods, startup costs, expert analysis, replication playbook, and more actionable insights.
Upgrade to PremiumInstant access to all founder journeys
Founders with similar journeys or strategies
I was frustrated with how hard it was to find real information about starting an online business. Everyone talked about raising venture capital, but I...
I built the first version of Gumroad in a weekend while working at Pinterest. The idea was simple: let anyone sell anything with just a link. The W...
I built Buttondown as a side project while working at Stripe. 5-10% month-over-month growth for years led to $392K/year. Side Project First I sc...
Get more founder journeys like this delivered to your inbox every week.