Ask HN: Should I give up and get a job?
I'm building a programmable PostgreSQL proxy in Rust. The idea is to make it easy to consume the replication stream so it can do query caching with automatic invalidation, and so that people can build custom partitioning, caching, or real-time features on top of PostgreSQL. The proxy part is implemented, but there's still a lot of work to add the replication and caching features, and to test and polish everything to production standards - databases are serious business. The project is on github here: https://github.com/riverdb/riverdb
I don't even know if this is something people would want or pay for if I completed it. And then there's the task of marketing/selling it, which is way outside of my skillset. Should I just give up (grow up?) and get a job? Is it worth pressing on here? I'm not consuming my savings, but neither am I making financial progress, and I'm not getting any younger.
246 comments
[ 4.4 ms ] story [ 262 ms ] threadWhen GCC came out the market disappeared for the cheap C compilers and even Microsoft got pulled into the discount market such as the MSDN subscription which entitles you to a lot of software and is an astonishingly good deal if you really use it.
It’s, I think, Visual Studio Subscription now (though the name has changed several times), but the licenses are for quite limited purposes, and designed to prevent the very high prices Microsoft charges end users for their software from being a barrier for vendors developing and supporting solutions that force end users to buy the software.
There is a huge untapped potential for selling software and services to businesses...but for some reason most developers never bother to even learn anything about the businesses they work in, let alone the businesses they don't.
It’s easier to target the equivalent of ‘spherical cows’ which don’t have pesky, irritating, problems like interop with 3 different major versions of windows (all old) or the equivalent.
I've bought a lot of licenses for developer tools - text editors, version control GUIs, a decompiler, multiple database clients, a client for SFTP, another one for testing HTTP APIs, multiple graphic editors, various ssh/shell tools for the phone... There's a lot of commercial dev tools out there and if you don't mind the fact that you will be dealing almost exclusively with tech dudes all the time it's a great opportunity.
Here is your problem. People give you money if you solve a problem for them (this is called a "value proposition"). You need to get into talks with people who need this enough to whip out the credit card. And you need many of those...
I do like your idea - but honestly, that is a very specific problem that few people have, and those who do are not necessarily the one to make buying decisions. And then there's the "why not just put all of this onto AWS, yolo" folks. Not to mention that your venture right now has a bus number of one: Even if I decided to put cash down on your product, what happens to the code should you get hit by a bus? (This is why it will be hard for you to get business customers)
You need either
- someone who can do the marketing - someone to shadow you - several high-profile customers who are willing to partly finance the development
or you need to understand this as a hobby project.
Also, don't underestimate the number of people with money who are just trying to make an excel spreadsheet work for their time forecasting, budget, team vacation time, building depreciation schedule, etc.
They are really struggling to make sense of a digital world and you can help. Whatever you do, it won't be giving up.
I wish there was a way for sales/marketing guys could team up with talented technical guys like you.
The usual advice here, although somewhat cliched, is get a cofounder, find a business partner or figure out a way to hire one.
They are all in this YC-Level community https://founderscafe.io
Go get a well paying job, and start a family. Stop chasing the startup dream, luck has a lot to do with it, and you didn't get lucky. But it's okay, there's no shame in that. You gave it your best shot, but there's more to life than making a startup. You can be extremely happy having a well paying job in tech and raising a wonderful family.
This is true.
> He gave up a job at Google and beats himself up every day for literally giving up millions, probably well over $10M had he just stayed.
This is also a biased success story. Very few people will experience Google-like returns.
This is why you're spending years and years spinning your wheels getting nowhere. It's a classic mistake pretty much everyone makes.
Very fundamental rule number one: Only build something AFTER you sell it FIRST.
The only time that rule doesn't apply is if you're doing a hobby project / labor of love and you're okay never making money off it.
EDIT: BTW as far as marketing and selling being outside your skillset, it's outside everyone's skillset at some point far enough back in their careers. i.e. nobody is simply born with it. The point is you develop those skills through pain, trial and error. That, or you find a business partner to handle it for you.
It's a common narrative that it's harder to be more successful as a solo founder. I don't agree because actually 66% of startups fail because of cofounder conflicts. Also startups with single founders tended to last longer and eventually achieve higher revenue (52% startups with succesful exits is solo founded). So if you are a solo founder, your risk of failing goes down significantly.
As for having someone to share your struggles, I joined this solo-ish founders community called https://FoundersCafe.io and they are bascially my co-founders without any of the drama.
I also hired out all the skillsets I lack.
Being a solo founder is fucking awesome as long as you have your needs met (emotional support + skill gaps)
I've come to understand this reluctantly over the past couple months of soul searching. I would not try to start coding again (with creating a business in mind) without having first found 5-10 customers.
Someone telling OP "interesting, I would buy it" might even be worse than not trying selling it at all, as it might fool OP into working even harder for nothing in the end.
And then you have to size the market. How many other people have that problem? How do you reach those people?
"hey I'm thinking of making this project, do you think you'd use it?" // "Yeah sure, seems helpful I guess"
you have:
"hey, person who told me they'd use this, here's the latest update where I've implemented X feature you and some others requested, I also am trialing Y feature that I think might be cool, and I polished Z item that I saw some others were getting hung up on, do you mind trying the latest version out and letting me know what you think?"
If you're unable/unwilling to maintain those kinds of relationships/conversations over the long term, consider a PM.
Not quite: saying you'd buy something and actually forking over the cash are two very different things. I've expressed interest in many a project that—when actually made available—I decided I didn't really need after all.
Is it always practical to actually sell a not-yet-existing product? No. But do you get the same data from just talking to people? Not at all.
So for example you can ask them this:
"Give me your honest answer. Let's say I had the product ready and finished and perfected right now... would you or would you not buy it from me?"
You really need to drive at this, you don't need to actually "sell" it. Validating an idea isn't a casual thing, but you don't have to go as far as actually selling it.
It is totally possible to get these real answers out of people. Your human, they're human and are able to simulate hypotheticals quite accurately if the simulation is framed correctly.
In fact I find it highly highly mistaken to think that it is fundamentally impossible for humans to answer that question.
The odds your product will hit a target that you have no idea if it even exists or not (or what it looks like, or what it’s shape is), is often lower than if you do.
Definitely not impossible of course!
But the odds are long enough already, it’s silly to not shorten them if you can.
But absolutely some deep questions about who you are building for are essential. There was a good post yesterday[1] ostensibly about open source, & closed source, but the conclusion was really about assessing not just purpose but audience.
[1] https://writing.kemitchell.com/2022/02/12/Devs-Use-Closed-So... https://news.ycombinator.com/item?id=30536825 (2 points, 14 hours, 1 comments)
I'm not too sure about _absolutely_ selling before making, but definitely do a market research and check if you can find people who would, at least theoretically, buy the product you intend to make.
If you make before you sell, you're saying that you are a visionary and people cannot understand your genius. Those are not odds to my liking, but if you like them you do you.
It's a common narrative that it's harder to be more successful as a solo founder. I don't agree because actually 66% of startups fail because of cofounder conflicts. Also startups with single founders tended to last longer and eventually achieve higher revenue (52% startups with succesful exits is solo founded). So if you are a solo founder, your risk of failing goes down significantly.
As for having someone to share your struggles, I joined this solo-ish founders community called https://FoundersCafe.io and they are bascially my co-founders without any of the drama.
I also hired out all the skillsets I lack.
Being a solo founder is fucking awesome as long as you have your needs met (emotional support + skill gaps)
Most often than not, they drag you down when incentives don't align.
This isn't fundamental. It's a single effective strategy out of many possible strategies. Think about it. Only certain people will buy something before you can show it to them and there's also a whole class of people who will only buy after they've seen the product.
This usually applies to random apps but it does not generalize to every class of software, for a few different reasons. Strategies need to be appropriate for the specific market.
You'll always need to sell, but selling first is a poor move for some products.
You seem to be creating solutions and looking for a problem. After 20 years you ought to know where the painpoints are in software, build something to fix those. That's what people will pay for.
Yes.
One other thing: taking a break is not the same as giving up. I've tried and failed at multiple startups, taking jobs in between to rebuild my savings. There is always opportunity cost and risk involved in all that we do, you just have to learn to be comfortable with whatever decision you make.
From what I read, it sounds way too "in the weeds" to really be a marketable good/service.
Otherwise, kids are rough. I'm not much older than you. I realize probably every day past 30 (give or take) I'm moving slower, and kids can be outright chaos demons. There's also the fact of wanting to be established with a corporation to some extent before you ask for time off to have a kid. There's a HUGE amount of corporate knowledge and wisdom to know, but the minimal is you may be outright unable to take time off during any probationary period at a new job. The probation period could be anywhere from 0 days (unlikely), 90 days (common), to a year (not unheard of). If your "have a kid" deadline is 2 years from now and you probably need to have a decent professional relationship with an employer for a year before having the kid, the time to get hired is in the next year. I think that means it's time to get serious with applying for work as a search can take considerable time. Probably time to get "desperate" about finding work around 9 months from now, as a guestimate?
Then you'll have the explaining problem partially solved. "I don't understand it either but I know to listen when ol' @eloff talks about Postgres performance."
It depends on the pay, current expenses, etc. For me, it would take 1-2 years on average (US tech job paying base close to $200k).
> Is there a target amount you computed? And how does it translate across living expenses, because I'm not living in Silicon Valley nor America.
This varies based on expenses and needs but generally, I would suggest at least 12 months of runway (for me this would be about $50-60k with some buffer, given that I support my family financially). This may take considerably longer for you if you're not in America.
> Is the prospect of not retiring in your 60s a possibility for you?
I am optimizing for financial security much sooner than 60. While I have an IRA account, it's not something I focus on or think about (or even actively invest into outside of when I have a tech job). I have a very high risk tolerance though, so take that with a grain of salt.
It's only my experience, but when I look back a lot of my effort was misdirected. It's very easy to feel like you're at 120% when you're focusing on what you know, but sometimes you need to shift the focus. I can't tell you whether the project you're working is worth continuing or whether there's money there, or whether you should continue it alone. I can tell you that I'm now working with 2 other founders and it's made me considerably happier than my solo efforts. There's always someone to talk to and whilst we still wear a lot of hats, we primarily concentrate on the bits we're good at and encourage each other to keep going.
I guess I'm saying - take a step back, maybe see if you can pull in help. You don't need to be happy all the time, but you do need to be content in your day to day (it's always going to be a struggle if you don't want to get up in the morning). If you're not there step back further and maybe see how a 'career' job works for you. Nothing is permanent, you can always step sideways, leave, or try something else until you figure out what works for you. Good luck!
Now that's not always the right choice for everyone, and there have been ups and downs, periods with too much work and not enough work, but overall it's been a good approach for me. One big thing is that since my finances are volatile I was very conservative in my purchases so that I could sleep at night knowing that I could cover all my debt and have a decent runway if the worst did happen and I had a hard time finding work for a long period of time.
If I would do it again I would be very tempted to just grind leet code as hard as possible and apply to a FANG company as it's going to be really hard to beat the total compensation from one of those companies even as a contractor and there is a huge amount of stability.
I can't predict the future but we're obviously in a turbulent time right now, if you're looking to start a family I would probably weigh stability pretty heavily right now.
Plus everyone expects to pull the goalie and you’ll have instant luck. My wife and are were both healthy and in our late twenties when we started trying. 3+ years of that and a round of IVF later and I finally have my first baby at 33.
Consider your top 3 assumptions relating to the following risks: users (do people want this?), business (can you sell it will it be profitable for you?), technical (can you complete the project in a way that will be useful for others?). These are the things that, if they're wrong, will mean you'll fail.
Then give yourself two days and find out the answers to those questions.
The only way to find that out in that time is to go out and talk to people. Give them a call; email them; reach out on Twitter - just get that feedback however you can. Be open to learning and being wrong, even though you've worked on this for a long time.
Some suggestions:
Users - find some people you think would want this and talk to them about it.
Business - find someone who's sold something similar successfully and talk through what they did with them.
Technical - speak with people who have written similar projects at the database layer.
Engineers don't typically like picking up the phone, but there's no alternative to asking questions and knocking down those risks as quickly as you can.
Problem 2: can that problem be solved with a product?
Problem 3: can that product be built?
Reading up on product management has helped me a lot here. Empowered (Marty Cagan) is good.
Get a job and switch to making it your hobby/side-hussle/itch-to-scratch project while $company pays your mortgage etc. Doing it that way would take a lot of the 'do I / Don't I' pressure off you.
If, after polishing it up a bit in your evenings/weeks-ends it still looks promising then yeah, that is the time to decide.
If you get good PR submitters, use them to help out and build up a relationship with them... you never know when you might need a good person like that as either a first-hire or a co-investor.
I read somewhere that the retention rate for SV IT techies tanks after about 3 years.... think about that and view the option of "taking a day job" not as a defeat but rather in the way of "hey, I've just got my own personal seed-funder for the next three years or so".
Best case - You get to hand your notice in after 6 months. Worst case? You polished your project and nurtured it for 3 years .... is it viable or not?
> I don't even know if this is something people would want or pay for if I completed it.
These two statements are incompatible. If they're true, this isn't a business, it's a hobby. That's not to say it's bad thing — I personally rewrote Redis on DynamoDB in Go over the course of the pandemic, and I learned a ton. But not a single person will ever pay for it.
I also wrote a scheduling system for the folks who run my gym as well, and they take me to lunch every chance they get to ask for new features and beg me to take more of their money.
The difference is classic YC startup lore, you make something people want. Not hope that people will want what you make. The former is a business, the latter is a hobby. Both are equally valid based on your life situation, but it doesn't work to confuse the two.
Yes, you should give up and get a full-time job. I'm a developer and I've founded 3 startups over the past 10 years. None made it to unicorn status, but they all passed the $2M in annual revenue mark.
Get a full-time job and make real money as an engineer. While you are doing that you can work on meeting the right co-founder and finding the right thing to build. The right product and market will pull you away from your job, you won't have to try and force it like you have been.
Your recent venture in profile is hard to put into the 'sell first, then build' bin. That wasn't your exact suggestion though - as I understand you, you say to have a passion for selling your product.
This said, I feel like you could map out the 'how I knew I could sell this at a price point viable for me and my customers.' Mind digging into that a bit?
Background: I'm getting better at my products/companies, but I've been burned recently taking on entirely too much (doing things that will actually never scale) because I was paid low thousands per setup and $500/month (I thought was huge for a SaaS), but that setup never got under 1-2 months like I wanted.
I've killed 5+ startups in the first few months though by prioritizing talking to potential customers over building software. Hard to do as an introvert that loves building things.
My recommendation to get a full-time job comes from the belief that if you are so focused on coming up with a winning idea, it becomes more illusive. (See: http://paulgraham.com/startupideas.html)
* Running Lean by Ash Maurya
* Business Model Generation by Alexander Osterwalder
When it comes to polishing your product, the growth.design course is excellent.
Sell before you build a full fledged product. Create traction.
If it's not making any money yet, it's a side project. I plan to take at least a couple of months off and work on these full time, but they're still just side projects .
Otherwise, you're basically gambling. The odds of being able to develop software solo that turns into something profitable, are not in your favor.
The idea is solid, but say you're the CTO of a tech startup. Would you trust this software with handling your data. Even if your intentions are good, I strongly doubt you have the money to hire a QA team to make sure you're not accidentally dropping data all over the place.
Even if you give it out for free, it poses an unacceptable level of risk. My side projects are almost entirely video games, the worst that happens is it crashes and you have to restart the game. I download random games from Itch.io all the time, as an individual I can take the risk on a strange binary breaking my Windows install.
I would never do this in a corporate environment.
I hope I wasn't too harsh, but this is an extremely difficult field to break into since the barrier to entry is so high. Realistically, you would need a small team.
Alternatively, let's just say you create another product. If you can prove your solution as a component of that, then maybe you'll be able to convince others to take the risk.