Ask HN: How did you choose an idea to work on?
16 comments
I've been trying to do this for a few years (admittedly failing).
You can come up with complicated models as some do, but I prefer simple models (3 dimensions or less): What do you want? What are you willing to build/iterate on for years? What can you find feedback on fast?
(1) because if you're using it you'll build something nice and have a lot of insight internally.
(2) because you'll need to work hard, fast, and for a long time to make something great.
(3) your reward systems depend on feedback.
Sometimes answering these questions isn't straightforward so I recommend building a lot of smallish things without an outcome in mind before answers to these questions become clear. You tend to know when you're well suited to something because you can't stop yourself from coming up with ideas and working on it.
Example: I spent a long time building/understanding http://choremate.co because it's something I NEEDED. Despite finding insight, (1) I found the work hard to do because it was boring (I don't enjoy CRUDing web apps), (2) I'm not innately interested in the research domain (behavioral and institutional economics), (3) feedback is best found living in a large shared household like a coop or fraternity and I don't live in that situation.
You can come up with complicated models as some do, but I prefer simple models (3 dimensions or less): What do you want? What are you willing to build/iterate on for years? What can you find feedback on fast?
(1) because if you're using it you'll build something nice and have a lot of insight internally.
(2) because you'll need to work hard, fast, and for a long time to make something great.
(3) your reward systems depend on feedback.
Sometimes answering these questions isn't straightforward so I recommend building a lot of smallish things without an outcome in mind before answers to these questions become clear. You tend to know when you're well suited to something because you can't stop yourself from coming up with ideas and working on it.
Example: I spent a long time building/understanding http://choremate.co because it's something I NEEDED. Despite finding insight, (1) I found the work hard to do because it was boring (I don't enjoy CRUDing web apps), (2) I'm not innately interested in the research domain (behavioral and institutional economics), (3) feedback is best found living in a large shared household like a coop or fraternity and I don't live in that situation.
Let me start with the most obvious one. Start working on a project that you actually need and are going to use, preferably in long term. That is how I started Bilinual Project around 2 years ago: http://www.bilinual.com
They aren't all good ideas. Start with your goals.
Do you want to "build something" where you follow through all the way to launch and hit some # of users that will feel like you "did it"?
Do you want to replace your day job? Do you want to raise funding and explode into the sun? Do you want to learn specific things throughout the process?
After you know why you want to work something, then rank your possible projects through the lens of "most likely to achieve my goals".
Lots to consider during that process: how excited you are about the thing, dogfooding opportunities, market size, competitive landscape, problem scale, technical challenges, business overhead (admin required beyond just coding the product).
Without any additional context I'd suggest building something that: you'll use (helps with motivation and you are your first user), will stretch your skills and challenge you to improve weaknesses (even if it fails, you've grown), can be launched then improved and extended through incremental iteration (avoid an all-or-nothing megaproject).
Do you want to "build something" where you follow through all the way to launch and hit some # of users that will feel like you "did it"?
Do you want to replace your day job? Do you want to raise funding and explode into the sun? Do you want to learn specific things throughout the process?
After you know why you want to work something, then rank your possible projects through the lens of "most likely to achieve my goals".
Lots to consider during that process: how excited you are about the thing, dogfooding opportunities, market size, competitive landscape, problem scale, technical challenges, business overhead (admin required beyond just coding the product).
Without any additional context I'd suggest building something that: you'll use (helps with motivation and you are your first user), will stretch your skills and challenge you to improve weaknesses (even if it fails, you've grown), can be launched then improved and extended through incremental iteration (avoid an all-or-nothing megaproject).
Thank you for your take - it's helpful to hear your approach
"Start with your goals"
Very well said and I am amazed to see that a lot of us fail to realize this key thing about life. If you don't have goals of some sort, it becomes a lot harder to do anything with a purpose. I am not saying everyone needs to have some crazy goals but you need to ask yourself: where do I see myself in next 2,5,7,10 years and you work your way through it then.
Starting a side project also should follow this. Yes we all have lot of ideas that are similar, nothing new. We all struggle to focus on that ONE thing to start with. The reason is because we are fearful. We fear what if that ONE thing fails since we have so many ideas. We analyze and paralyze. Result is: we do nothing.
Ok so lets say you have your 2,5 years goal set. Now lets say multiple ideas could take you there in theory (since you don't know for sure without trying execeution). In that case, give yourself few days.
Day 1: Make a list of all your best ideas that you want to execute on.
Day 2: You have a gun to your head. Take 1 idea off that list. Only one. leave the rest. Once you discard the idea, you can not add it back to the list.
Day 3-n: Repeat Day 2 until you are left with that 1 idea.
Final day: execute, execute, execute. No remorse. No guilt on dropping other ideas.
Very well said and I am amazed to see that a lot of us fail to realize this key thing about life. If you don't have goals of some sort, it becomes a lot harder to do anything with a purpose. I am not saying everyone needs to have some crazy goals but you need to ask yourself: where do I see myself in next 2,5,7,10 years and you work your way through it then.
Starting a side project also should follow this. Yes we all have lot of ideas that are similar, nothing new. We all struggle to focus on that ONE thing to start with. The reason is because we are fearful. We fear what if that ONE thing fails since we have so many ideas. We analyze and paralyze. Result is: we do nothing.
Ok so lets say you have your 2,5 years goal set. Now lets say multiple ideas could take you there in theory (since you don't know for sure without trying execeution). In that case, give yourself few days.
Day 1: Make a list of all your best ideas that you want to execute on.
Day 2: You have a gun to your head. Take 1 idea off that list. Only one. leave the rest. Once you discard the idea, you can not add it back to the list.
Day 3-n: Repeat Day 2 until you are left with that 1 idea.
Final day: execute, execute, execute. No remorse. No guilt on dropping other ideas.
I wish I had a good answer for you. All I can really offer is commiseration, by way of saying that I have the same issue. I spread myself way too thin, trying to work on multiple projects at the same time, instead of picking on and working on it exclusively (or even near exclusively).
That said, to the extent that I have gotten any better at this over the years, I did the following:
I created a page on my personal wiki, with a "todo list". At the very top, before any of the todo items are a series of questions. Every time I look at the list, I'm reminded to ask myself those questions.
I don't have the list in front of me right this second, and I'm not sure I'd want to share the exact list anyway, as it's kinda subjective and personal. But assume it includes things like:
1. What is the single most important thing I could be doing RIGHT NOW, vis-a-vis creating revenue?
2. Whatever I'm about to do, WHY am I doing it NOW?
etc.
It helps a little bit, but not completely since so much of what I do is kept in my head and I don't consult this page all that often. But at least I get a little reminder every now and then of what I'm supposed to be focusing on, in terms of the strategy I wrote down.
That said, to the extent that I have gotten any better at this over the years, I did the following:
I created a page on my personal wiki, with a "todo list". At the very top, before any of the todo items are a series of questions. Every time I look at the list, I'm reminded to ask myself those questions.
I don't have the list in front of me right this second, and I'm not sure I'd want to share the exact list anyway, as it's kinda subjective and personal. But assume it includes things like:
1. What is the single most important thing I could be doing RIGHT NOW, vis-a-vis creating revenue?
2. Whatever I'm about to do, WHY am I doing it NOW?
etc.
It helps a little bit, but not completely since so much of what I do is kept in my head and I don't consult this page all that often. But at least I get a little reminder every now and then of what I'm supposed to be focusing on, in terms of the strategy I wrote down.
Thank you, I appreciate your perspective. Sounds like you've worked on this problem yourself by working at it from a different direction, namely starting from what you want to do/why you want it and moving back towards the thing to be doing.
Sort of. I have several projects I'm working on, and in a sense I don't "know" which one to focus on. But underneath everything, I know that I ultimately want to build a company that makes money. Not that "make money" is the only thing I want to do, but it matters for $REASONS.
Anyway, given that, I can work backwards from "what action(s) are most likely to lead to somebody paying me even 1 dollar for any of this?" And if I keep that "time to first dollar" in mind as a north-star of sorts, it helps me focus. To a degree.
In reality I don't always actually know what is going to minimize "time to first dollar", and sometimes I just say "screw it" and work on what seems like the most fun. So it's a really rough sort of heuristic at best. This is probably one reason why we aren't making any money yet. But that's OK, because other goals besides "make money" include "learn cool new stuff" and "have fun."
Anyway, given that, I can work backwards from "what action(s) are most likely to lead to somebody paying me even 1 dollar for any of this?" And if I keep that "time to first dollar" in mind as a north-star of sorts, it helps me focus. To a degree.
In reality I don't always actually know what is going to minimize "time to first dollar", and sometimes I just say "screw it" and work on what seems like the most fun. So it's a really rough sort of heuristic at best. This is probably one reason why we aren't making any money yet. But that's OK, because other goals besides "make money" include "learn cool new stuff" and "have fun."
Over the years I've come to realize that the answers to your question is like asking what the meaning of life is, it's very subjective. So instead of giving you practical advice like "Write an essay on why not to make and idea." (which is actually really effective by the way when done right), I will give you some indirect advice:
1. Know the difference between reasoning and logic. Reasoning is like telling a story whereas logic is more like writing an axiomatic proof. Don't give in to reason, only logic will help you find the right answer. I call reasoning "scheming" when it's done in the context that the person doing it believes they are thinking logically but are just fooling themselves.
2. Shut the hell up! Stop your inner narrative in your life keeping you stuck in life and just observe the world around you. I find it helpful when I’m _doing_ something and not making progress to _not-do_ what I’m doing—it’s quite magical and isn’t what it sounds like. What I mean is you assume an indifferent mental state which frees up your brain's resources to take notice and inventory of details you and others missed about a problem.
3. Follow the path with heart! Is there a splinter in your mind but can't quite put your finger on it? Don't expect to find your life's purpose instantly, it takes time. The best advice I can give you is pursue what makes you the most curious, not what will make the most money. This is because things you are curious about take less energy to think and learn about.
4. Be compassionate, not passionate. What I mean by that is don't pursue "ideas" but pursue what makes you happy and others happy; be a buddha and end the cycle of suffering. All that matters in this world is happiness. If making money like Warren Buffett does makes you truly happy (but not for the money itself) then become a stock investor first, programmer second--you see what I'm saying? Too many programmers have paralysis when it comes to side project ideas because what they enjoy most is programming and they are already programmers, so shouldn't you be happy by now ;)
5. Be patient! Don't try to think of an idea, rather find a problem (which could just be a tedious task you do often) as a consumer, not a programmer, that really bothers you. Then, rather than make a complete product at first, do experiments first not by building a product but testing assumptions or hypotheses, doing a trial, and draw conclusions from what you learned. Over time, all those experiments will result in _learning_ which is the modus operandi of business.
6. Do your homework! Spend a good year at least researching all you can about the problem you are solving (unless you're just making a metoo app) and why it isn't solved yet, because you may find out for example the problem is intentional which means it can't be fixed, it is corrupt by design. You may discover a simpler way of solving the problem that incumbents were too fancy to notice. Treat your problem like it's a real conspiracy which forces you to think differently about the problem than others; rather than just being an engineering challenge, it becomes web of even deeper problems, and you may just find that your problem is just the tip of the iceberg.
1. Know the difference between reasoning and logic. Reasoning is like telling a story whereas logic is more like writing an axiomatic proof. Don't give in to reason, only logic will help you find the right answer. I call reasoning "scheming" when it's done in the context that the person doing it believes they are thinking logically but are just fooling themselves.
2. Shut the hell up! Stop your inner narrative in your life keeping you stuck in life and just observe the world around you. I find it helpful when I’m _doing_ something and not making progress to _not-do_ what I’m doing—it’s quite magical and isn’t what it sounds like. What I mean is you assume an indifferent mental state which frees up your brain's resources to take notice and inventory of details you and others missed about a problem.
3. Follow the path with heart! Is there a splinter in your mind but can't quite put your finger on it? Don't expect to find your life's purpose instantly, it takes time. The best advice I can give you is pursue what makes you the most curious, not what will make the most money. This is because things you are curious about take less energy to think and learn about.
4. Be compassionate, not passionate. What I mean by that is don't pursue "ideas" but pursue what makes you happy and others happy; be a buddha and end the cycle of suffering. All that matters in this world is happiness. If making money like Warren Buffett does makes you truly happy (but not for the money itself) then become a stock investor first, programmer second--you see what I'm saying? Too many programmers have paralysis when it comes to side project ideas because what they enjoy most is programming and they are already programmers, so shouldn't you be happy by now ;)
5. Be patient! Don't try to think of an idea, rather find a problem (which could just be a tedious task you do often) as a consumer, not a programmer, that really bothers you. Then, rather than make a complete product at first, do experiments first not by building a product but testing assumptions or hypotheses, doing a trial, and draw conclusions from what you learned. Over time, all those experiments will result in _learning_ which is the modus operandi of business.
6. Do your homework! Spend a good year at least researching all you can about the problem you are solving (unless you're just making a metoo app) and why it isn't solved yet, because you may find out for example the problem is intentional which means it can't be fixed, it is corrupt by design. You may discover a simpler way of solving the problem that incumbents were too fancy to notice. Treat your problem like it's a real conspiracy which forces you to think differently about the problem than others; rather than just being an engineering challenge, it becomes web of even deeper problems, and you may just find that your problem is just the tip of the iceberg.
Unfortunately, I don't have a clear cut answer for you either, but share my experience:
1. I didn't "choose" the idea, the idea chose me. i.e. It took on a life of its own, and we had product market fit.
2. The fact that you are dabbling might actually be a good thing i.e. the explore-exploit dilemma.
1. I didn't "choose" the idea, the idea chose me. i.e. It took on a life of its own, and we had product market fit.
2. The fact that you are dabbling might actually be a good thing i.e. the explore-exploit dilemma.
Just determine the reason why you're working on these ideas.
Is it for fun? Education? Money?
You should be able to determine which of these ideas are more aligned with your end goal.
Is it for fun? Education? Money?
You should be able to determine which of these ideas are more aligned with your end goal.
Select the top 3 ideas. Put out landing pages for each with mock demos. Don't write code yet. Gently engage people online (ex: twitter) to gauge for concrete interest. You'll naturally start focusing on the idea with the most excitement and engagement.
Bonus point if you get people to commit to the idea with money.
Bonus point if you get people to commit to the idea with money.
Honestly OP this is your best solution. First find out if there's an easy accessible market for your project, and if it had costumes interested in it.
I promise you that is better to start off with validation of your concept, and an interested market (who you can then start testing your iterations with), before you write a line of code.
If you can't get product market fit, then either there is no market, or you've simply not found the market yet/sold the concept of your product well/or the market just doesn't want what your offering.
I promise you that is better to start off with validation of your concept, and an interested market (who you can then start testing your iterations with), before you write a line of code.
If you can't get product market fit, then either there is no market, or you've simply not found the market yet/sold the concept of your product well/or the market just doesn't want what your offering.
Most ideas will just happen without me. Those ideas are either not worth doing, because someone else can do the work and I can still reap some of the benefits. Or they might result in more competition.
So the way I pick, is the hardest idea I can carry. The one that is least likely to happen without me. This also tends to be the most valuable, because I wouldn't even be pondering doing it if it weren't worth it. It ends up being the easiest idea to pitch for funding, and the easiest kind of company to sell, if that's the goal. It's the one that some tech giant like Facebook or Tencent can't just come in and throw a million dollars at to do- and as such they might even throw a million dollars at you to buy it. Sometimes current revenue doesn't even really matter; something like YouTube is worthless on its own, but plugged into the behemoth that is Google Ads, the revenue gushes in.
The next step I do is work on it every day. As in actual work, not just planning. This filters out some ideas; I can't reform the education system or work on nuclear fusion one day at a time, but there are things I can do.
So the way I pick, is the hardest idea I can carry. The one that is least likely to happen without me. This also tends to be the most valuable, because I wouldn't even be pondering doing it if it weren't worth it. It ends up being the easiest idea to pitch for funding, and the easiest kind of company to sell, if that's the goal. It's the one that some tech giant like Facebook or Tencent can't just come in and throw a million dollars at to do- and as such they might even throw a million dollars at you to buy it. Sometimes current revenue doesn't even really matter; something like YouTube is worthless on its own, but plugged into the behemoth that is Google Ads, the revenue gushes in.
The next step I do is work on it every day. As in actual work, not just planning. This filters out some ideas; I can't reform the education system or work on nuclear fusion one day at a time, but there are things I can do.
> So the way I pick, is the hardest idea I can carry. The one that is least likely to happen without me.
This was a really useful metric. Thanks.
This was a really useful metric. Thanks.
> I have a ton of ideas and no issues executing them
Branch out to ideas that skirt the edges of your knowledge. If you're too comfortable you're not growing stronger.
> They're all generally good ideas, and with dedication each one could become something useful
Then they need to battle. Think evolutionary. Find a concrete set of resources for them to fight over (money, time, space, popularity?) and score them. Iterate them. Combine them. Kill them. May the best idea win!
If that doesn't work then find the most profitable and productive idea that can serve the most people. Those tend to be few and far between.
Branch out to ideas that skirt the edges of your knowledge. If you're too comfortable you're not growing stronger.
> They're all generally good ideas, and with dedication each one could become something useful
Then they need to battle. Think evolutionary. Find a concrete set of resources for them to fight over (money, time, space, popularity?) and score them. Iterate them. Combine them. Kill them. May the best idea win!
If that doesn't work then find the most profitable and productive idea that can serve the most people. Those tend to be few and far between.
Does anyone have experience with this? Any methods on how to choose an idea to pursue? Maybe specific criteria to evaluate each one on? I would appreciate any guidance or advice on this.