← All distillations
book·one-page markdown

Anything You Want

Derek Sivers

Operating rules for building a company you actually want to own, drawn from Derek Sivers running CD Baby from a favor for friends to a $22 million sale without investors, a plan, or a vision statement.

Chat with the guideChat with this corpus

This governs what a founder decides about their own company: what to promise, who to serve, what to charge, how much to grow, and what job to keep for themselves. Sivers treats a business as a personal creation rather than an obligation to maximize anything, so the right size and the right customer are the ones the owner actually wants. The evidence is CD Baby, which started as a favor, cost about $500, and was profitable in its second month.

Do

  • Write the mission as promises to the customer, not as a growth plan. CD Baby's entire plan was four commitments: pay artists weekly, hand over the full contact details of every buyer because those are the artist's fans, never drop an artist for low sales, never sell placement. Everything else got decided against those four. [Source: "Just Selling My CD"]
  • Make the pricing simple enough to read at a glance. He copied a local record store's consignment terms (seller sets the price, flat cut per sale, weekly payout) and added a flat setup fee for his own time. A $35 setup fee and a $4 cut funded the company for years. [Source: "Just Selling My CD"]
  • Start today at a tiny fraction of the eventual scale. Teach one paying student instead of building a school. Charter one plane instead of founding an airline. CD Baby's entire first year was a short list of CDs where each buy button just emailed him the order. [Source: "Start Now, No Funding Needed"]
  • Share what you already have access to, and charge only once it stops being a favor. Before CD Baby he had scanned copyright forms onto his site, written free trademark instructions, and lent musician friends his barcode account and servers. Each one became real revenue without ever having been conceived as a business. [Source: "The Co-op Business Model"]
  • Take many small customers over one large one. A single big client effectively becomes your boss, since their mood and their continued business set your decisions. With thousands of small ones, losing any single customer barely registers. [Source: "Proudly Exclude People"]
  • Staff customer service like it is the product, because for you it may be. Musicians told him they chose CD Baby because a real person picked up the phone and answered email, not because of price or features. He put 28 of 85 employees into customer service and treated it as a profit center rather than a cost. [Source: "Customer Service Is Everything"]
  • Answer each new question by teaching the reasoning, then write it down and hand over the decision. Three years in he was working from morning to night and still being interrupted constantly. Within two months of explaining the philosophy behind each answer and putting it in a shared manual, the questions stopped, and the company grew from roughly $1 million to $20 million without him in the room. [Source: "Delegate or Die"]
  • Build every process to carry twice the current load. The company roughly doubled in size six years running. A business visibly straining under its growth signals it cannot take more, while one with headroom signals it can. [Source: "Delegate or Die"]
  • Cut every unnecessary word before sending anything at scale. Writing to roughly two million customers taught him that one ambiguous sentence produces tens of thousands of confused replies and a real support bill. Unclear website copy fails silently instead, which is why most people never learn this. [Source: "Little Things Make All the Difference"]
  • Keep verifying the one job that would be fatal to lose. He trained someone carefully to deliver every artist's music to every digital retailer weekly, then trusted him completely, and found out months later that deliveries to several major retailers had quietly stopped. He ran them himself for six months to rebuild the system. [Source: "Make It Anything You Want"]

Don't

  • Don't keep pushing something that is not landing. He spent roughly a decade promoting music projects through relentless pitching and always made progress, but only through force. CD Baby grew with almost no promotion. Persistence means improving and inventing new versions, not pushing the same one harder. [Source: "Hell Yeah or No"]
  • Don't defend your original plan against what customers are actually doing. He built CD Baby as a credit card tool that pointed back to musicians' own sites. A confused customer email asking about new releases showed him buyers already thought it was a store, so he made it one. [Source: "Hell Yeah or No"]
  • Don't size a market by taking a small percentage of a big number. A musician pressed 10,000 CDs and rented mailing infrastructure, certain that one percent of a magazine's million readers would clear his stock. The ad ran and he sold four copies. [Source: "Proudly Exclude People"]
  • Don't add your two cents to finished work. A small offhand suggestion from the owner functions as an order rather than a note, and it quietly takes away the employee's ownership of their own project. Say something only when something is substantially wrong. [Source: "Delegate or Die"]
  • Don't hand over responsibility along with authority. After giving employees open-ended power over office decisions, health plans and profit sharing, he found they had redirected company profits to themselves. Reversing it made him deeply unpopular and ended with him working remotely rather than repairing the relationship with an 85 person staff. [Source: "Make It Anything You Want"]
  • Don't sign paperwork you have not read to save a filing fee. Years earlier he had signed his father's company's paperwork while raising money for recording gear, then took a bank teller's suggestion to fold CD Baby's finances into that same entity. It turned out to own 90% of his company, and buying majority control back cost $3.3 million. [Source: "It's About Being, Not Having"]
  • Don't write a permanent policy in response to one bad incident. A school banned grape juice after a spill, then orange juice after another, until only water was left. Airport shoe removal outlived the single event that caused it. [Source: "Customer Service Is Everything"]
  • Don't buy formalities sold to you through fear. No terms of service, no privacy policy, no lawyer on retainer, even at fifty employees with vendors pitching compliance packages. Plenty of ordinary small businesses run fine without them. [Source: "The Co-op Business Model"]
  • Don't assume growth past your comfort is progress. He tried to hold CD Baby at twenty employees, then again at fifty, and lost both times to demand. His own account is that the stories got less happy as the company got bigger, and that he was happiest running it nearly alone. [Source: "Make It Anything You Want"]

Hold

  • A company is a playground for your curiosity before it is a wealth vehicle. The closing test he gives is whether the business reflects what actually makes its owner happy, even if that means staying deliberately small. Held, small stops reading as failure. Not held, every decision defaults to whichever option is bigger. [Source: "Why You Need Your Own Company"]
  • You do not owe your company its own survival. Asked at a conference how he would stop musicians from eventually bypassing CD Baby, he said he did not care whether CD Baby survived, only whether musicians were served well. His warning is that companies quietly protect the problem they exist to solve, because solving it fully would end them. [Source: "You Can Afford to Be Generous"]
  • Doing it yourself slowly can be the whole point. He spent about fifteen years and a thousand shows becoming a competent singer against repeated advice to hire one, and kept doing all of CD Baby's programming himself while employees objected that features were shipping too slowly. He took the cost on purpose, because the point was becoming something rather than finishing something. [Source: "It's About Being, Not Having"]
  • You choose the yardstick you are graded by, or you inherit one. People measure themselves by net worth, by money given away, by lives touched deeply or broadly. His own is how many useful things he creates, which is why he could answer MBA visitors asking for growth rates and retained earnings with almost nothing. [Source: "Just One of Many Options"]

The one line

Build the smallest working version today, promise your customer something specific, serve many small buyers rather than one big one, and stop growing at the size you actually enjoy owning.