3 ms·
You should not start a startup. I like to think of it differently. It's semantics but I find that it aligns my thinking better. You should start a business if:
by blizkreeg 9y ago
You should not start a startup. I like to think of it differently. It's semantics but I find that it aligns my thinking better. You should start a business if:
- you are strongly driven _by the idea_ of doing business, i.e., at some point in your evolution, you become Walt Whitman. A business, which
- solves a problem, makes a product, or offers a service, that
- you think can you do better and make money while doing so, and
- you would rather face the uncertainty of it not working out than be employed working on a thing you are not motivated by.
You should not start a startup if things such as exits, raising VC, optics of running a startup, or some such affair is of primary concern to you. They should be in the service of running the business and serving its needs, not the other way around.
This is not a script that applies to just some lifestyle, sustainable businesses, in the parlance of SV. This is the same script that even the largest, most profitable, and most impactful companies (cue the disruption adjectives) on the planet have followed.
- CryptoFascist 9y agoThis! Especially the 'solves a problem or offers a service'. Any company without a viable plan to create surplus is a failure waiting to happen. It's why most startups fail... their purpose is to be a startup first, create a surplus if they luck out, or be flipped into a bubble market.
- timr 9y agoThe key word there is "viable plan", and you don't know that you have one until you do it. Many of the biggest, most successful companies are pivots from some lame initial product. There are plenty of bozos who don't try to solve a problem or offer a service, but there are plenty of dead startups who did all of that and still failed.