Starting a Startup
When you start something new you don't know what you don't know. If it's your first time building a startup there are some basic things that go against common knowledge or might be the opposite of what you'd usually do at a regular job. If you're already in a startup hub like Silicon Valley or New York City, then you can probably find a mentor and leverage the startup culture there. If you're in a city where software engineers are uncommon and tech startups are rare, having advice on how to start might help.
My education is from YC's online startup school, a bunch of books, and going out and doing it. I got through most of the initial development stage and stopped short of launching. The advice below is for early phase development since that's what I worked through.
Everything but the iffy advice on the bottom is good and should apply to all startups. This is a 101 for how to start a business with details I wish I would've known when I started.
LLC or C-Corp?
Startups are usually filed legally as either LLCs or C-Corps. The basic rule for which to do is this:
Choose C-Corp when taking funding.
Choose LLC when bootstrapping.
Which State to File In?
Unless you're already a lawyer and know what you're doing, just stick to the basics. Delaware is the state that people usually use because they have a special court system that expedites business cases. VCs will usually want you to be in Delaware before giving you money. Moving states after you're already established can be a pain too, so just keep it simple.
The only exception to this would be if you want to start an anonymous company for a more taboo industry. In that case Nevada has and anonymous LLC where your name isn't publicly listed as owner.
How to File?
Legal companies automate the process of incorporating your business. They'll charge a bit for the work but it's not much and if you're successful it'll be much cheaper than having to redo it with a $300/hr lawyer because you fudged details when filing.
There are lots of companies but I'm just going to recommend what I know works:
Choose Clerky for a C-Corp.
Choose Incorporate.com for an LLC.
Find a Co-Founder
Co-Founder problems are the #1 killer of startups so pick carefully. Identify your strengths and weaknesses as a founder. Try to find someone that fills in your weak spots so together your founding team doesn't have skill gaps.
You should have history with the person you're teaming with so you know if you'll work well together.
Co-Founder sources (ranked best to worst):
- best friend
- co-worker
- acquaintance
- stranger
What Does Your Company Do?
Make a 1-2 sentence explanation for what your company does. It should be straight to the point. Memorize it so you can answer the chronic company question for friends, family, VCs, peers, and everybody else.
example: "BonNochi is a hotel booking service for travelers with disabilities. We book accessible hotel rooms and guarantee in-room accessibility features for your trip!"
Make Money
Bootstrapping helps here but your company has to make money. Maybe you can raise rounds for a bit but eventually you'll need to be in the black.
Pick something with good margins and reasonable overhead costs.
Validate First
Do UX research to understand what you'll need to build before doing development. It's easy to change ideas or designs when you haven't spent the last two months writing code.
You also want to solve actual problems that lots of people have. If you're not solving a problem nobody should use your app. If your target demographic is too small, you're not going to grow much.
Potential customers need to be sold on the idea when they're giving you feedback. Look for things like:
"When will your product be released?"
"Can we sign up for updates on your release?"
Test designs
Once you know what to build you'll probably want to test the app designs. You could do user testing after months of development or you could hack together some design mockups in a few days. One is cheap and easy to change, the other not so much.
Product Development Pipeline
figure out what to build -> design it -> spec it -> build it
Focus on One Thing
Starting out is hard. You'll probably start by dreaming big but you have to be as quick to market as possible. Take the most critical feature of your service and just build that. Good targets are the things that make the most profit, have the most social impact, or are the most requested during UX research.
Make this one thing 10x better than anybody else can do it.
Punt on the hard stuff.
Figure Out HATED Things
Still stuck on what to build? Try to figure out what people HATE.
During UX research did people have one thing that was horrible to deal with? Build that.
KPI == Business Health
KPI is startup speak for "Key Performance Indicator" which is mumbo jumbo speak for "a key metric that tracks your company health well". Pick an important metric and use it as a proxy for your company's health.
If you get big use multiple metrics.
Don't Overwork
What's better than doing caffeine-fueled, round the clock overtime work to kick out your killer startup?
Doing moderated amounts of work with a clear focus. Give yourself a 9-5 with occasional crunch times and keep your social life. If you can't blow off steam and enjoy yourself, you're going to burn out.
Personal Opinions (iffy advice):
Here's some other advice that I personally think is good but I can't say if it's 100% the right thing to do. I can expand on anything if anybody wants more (email me).
- customers are the best investors, they give you money without taking equity. Start there
- saliency bias is real. if you can contract a designer to beautify your website before showing to VCs, do it (big money at stake).
- work full-time on the startup, there are way too many things to do, it's really an all-or-nothing proposition for a major product
- worry about product development not competition, it's not a zero-sum game for a scrappy startup
- log what happened each week to identify trends and self-correct
- random contracting is rough because developer skills vary considerably and they don't share your product vision
- get an adviser 3-4 years ahead of you (their wounds are still fresh and they're probably still doing startup work)
- build a community that provides value to people, ask for help from the people there for testing or UX interviews (warm intro, product evangelists, satisfying to solve their problems, fun people to meet too!, weak network effect)
- know thy customer (where do they hang out?, why do they like you?, demographic trends, personality trends, etc.)
- never make deals that feel fishy (hires, funding, business partners)
- set company values up front, short-term goals as needed
- set a long-term revenue/customer/funding target on day one and kill the startup if it doesn't perform (avoid being a zombie startup even if you're making some money)
- plan growth by orders of magnitude, 0->1->10->100->1,000->10,000…, short-term goals change at each milestone, number of customers is a good metric for this
- internally pick one large, incumbent company in your industry and make them your target (creates a false equivalency that you're doing good because you're beating company X on Y thing, morale booster). Find their weak spots and try to beat them there
- think about the future and try to guess how your industry will have changed in 30-50 years, build that
- bootstrapped businesses should start niche and grow from there
Best of luck if you're going after something on your own!