1. Product at a startup

§Ask five founders what product means and you get five answers: the app, the roadmap, the design, the thing engineering builds, the thing marketing sells. All of them are pieces. None of them is the job.

§The job is integration. A startup is a machine for learning what to build, and every part of that machine produces signal. Engineering discovers that the feature everyone wanted takes three months. Sales hears the same objection four times in a week. Support gets an email that reveals a use case nobody imagined. The market shifts under a competitor's launch. Each of these is a data point about the product, and each arrives somewhere else.

§Product is the place where they meet.

§ 1.1The integration point#

§In a large company, integration is a process. Signals are logged, tagged, prioritized and reviewed, and the delay between a signal arriving and a decision changing is measured in quarters. A startup cannot afford that delay, and does not need the process, because the whole company fits in one room.

§What it needs instead is a person through whom everything passes. Not a gatekeeper who slows things down, but the opposite: the one person who hears the sales objection, the engineering estimate and the support email in the same afternoon, and can see that they are three views of the same problem.

§This is why the role cannot be led in isolation. A product owner who sits apart from sales calls and support inboxes, writing specifications from a plan, is optimizing the one thing a startup does not lack: the ability to write things down. They are starving the one thing it does lack, which is contact with reality.

§ 1.2Speed of information#

§Consider what a company of one to five people actually has. It has a hypothesis about a problem worth solving. It has one to a few people who can build, and tools that let one person build what used to take a team. It has a runway measured in months. It does not have customers in volume, data in volume, or the luxury of being wrong slowly.

§In that situation, the constraint is not talent and it is not money. It is how fast the company learns. Every day between a customer saying something important and the team changing course is a day of runway spent building the wrong thing. Speed of information is the scarce resource, and product is the discipline that manages it.

§Steve Blank's customer development framework, written twenty years ago, made the same point in different words: a startup is a temporary organization searching for a repeatable business model, and the search is a learning loop. What Blank did not have to say in 2005, and what matters now, is that the loop's speed is set by how quickly signal reaches the person who decides.

§ 1.3What product is not#

§It helps to say what the job is not, because the misreadings are expensive.

§Product is not engineering. Engineering decides how to build and how long it takes. Product decides whether to build it at all, and what it must do for whom. A founder who confuses the two ends up with a very well-built solution to a problem nobody confirmed.

§Product is not design. Design decides how the thing looks and behaves. Product decides what experience it is meant to produce and for which person. The next two chapters go deep on this, because at a startup the product owner usually holds design too.

§Product is not marketing. Marketing decides how the world hears about the thing. Product decides what the thing is. They share a lot: the same customer, the same words, the same promise. They are not the same job.

§And product is not a committee. Every one of those disciplines contributes to the decision. One person makes it.

§ 1.4The decision, restated#

§Everything in this chapter reduces to one sentence. Product at a startup is the discipline of turning everything the company learns into one decision about what to build next, made by one person who is answerable for it.

§The next chapter is about that person.