What is Nostr?
Karnage
npub1r0r…q9ac
2023-10-28 08:12:20

Crosspost: When is it okay to just quit?

Via reddit:

" I have been working on the side with my partner on an app since more than a year ago. I work a day job to pay off my student loans and lately, have been losing hope on the startup. At this point, I'm only in it because 1) of my co-founder and 2) I have already given it a year of my life. Everyday is a struggle to work on it. I joined my co-founder because I believed in the idea and thought it could be a hit.

Several iterations and several "pivots" later, we have still not launched the app. Pivoting too often and too much was a mistake. Even the current set of features of the app aren't really good enough.

My personal life has taken a hit, I'm not able to give my 100% to my day job. My personal relationships have suffered tremendously, and there's also a lot of "tech-debt" that has accumulated because of subpar quality of code by the freelancers. I'm just constantly fixing one bug after another and I have lost the energy because it hasnt been rewarding at all for me. I can't even sell my positions and exit.

I'm trying to gain different perspectives here and trying to find out what people have done when they found themselves in a situation similar to mine "

My thoughts ----

1. Beware of sunk cost fallacy. Just because you put a year into it, doesn't mean you're somehow going to make up for it by dedicating any more time.

2. It seems the problem is the "idea" stage. This is where a lot of "startups" fail. New ideas are hard to execute and you have to wonder why if it's such a great idea that nobody has done it that way. Chances are it's a bad idea.

If I were this person I would take my losses and get my life together, then learn from my mistakes hopefully and give it another go later (or not).

The problem with a lot of startups is that they are solving non-problems.

The cure--

First time founder should stick to something that is already somewhat validated. In other words, other similar things exist and try to solve the same problem. If you can identify at least 3 other profitable companies that are solving similar problems (or better yet, your exact problem you're trying to solve in a more novel way), then you have a validated market.

Try to bootstrap a solution. Give yourself an unreasonably short amount of time to ship a solution. Launch when you are 80% there. Have a plan B if things don't pan out.

A pivot in the idea stage before the product is even shipped usually means shiny object syndrome. This is where a "better idea" is begging for your time. It's a distractions.

It's good to ask yourself a few questions before even starting (btw I can't use colons here)
1. Who are my 3 competitors
2. Who are my customers? You should be able to give them a concrete title. Who are they and what do they do professionally? If the answer is "everyone" you need to start over.
3. What is my target market size - quantify it (chances are it's bigger than you think)
4. How will I distribute my product? (in fact, this should be your #1 question and priority).

Then, identify your product, who it is for and what problem it solves. " X is a Y for Z and it {solves this problem}"

If you get this far, now identify the absolute minimum product required to provide that solution. Often the product is not even necessary, or you can solve the product manually via a service. Do that service to validate the need for your product. Solve the problem manually for people before spending months building an automation.
Author Public Key
npub1r0rs5q2gk0e3dk3nlc7gnu378ec6cnlenqp8a3cjhyzu6f8k5sgs4sq9ac