Ask HN: Any pain points worth building an MVP for?
Software engineer here looking for ideas to build MVPs for bootstrapping his way into becoming an Indie Hacker. Any pain points worth building an MVP for?
44 comments
This is something I’d love to build but don’t have time. I was talking to a police officer a couple months ago and he was telling me the difficulties they have dealing with people that speak no English during emergency situations (someone’s bleeding out).
They have a language service they can use but it’s slow to loop another person in. They end up resorting to Google Translate which is also kinda slow to type in their questions. The voice to text feature is finicky too.
What he wants is a website / app with a series of pre recorded questions. The answers are simple yes / no nods. This could potentially save someone’s life.
I honestly think this should be a free website. But if you’re looking to start a business it could be a great way to build credibility in the space.
They have a language service they can use but it’s slow to loop another person in. They end up resorting to Google Translate which is also kinda slow to type in their questions. The voice to text feature is finicky too.
What he wants is a website / app with a series of pre recorded questions. The answers are simple yes / no nods. This could potentially save someone’s life.
I honestly think this should be a free website. But if you’re looking to start a business it could be a great way to build credibility in the space.
Would you make this purely for local law enforcement and first responders? What questions would be the highest priority for the use case?
Would you also have domain specific use cases as well?
Would you also have domain specific use cases as well?
Great questions, I’m not sure. If I were approaching this I’d try to partner with a police officer as the domain expert.
I imagine it wouldn’t be hard to find someone interested in helping - most people love talking about their problems. Maybe join some Facebook groups? Or get in touch with your local precinct?
I imagine it wouldn’t be hard to find someone interested in helping - most people love talking about their problems. Maybe join some Facebook groups? Or get in touch with your local precinct?
I have some Nginx access logs. Not many logs. I need to parse those logs and analyse them.
Some questions I want to answer from my logs:
1. Which API is slow
2. Which API is broken, ie. returning 4xx or 5xx
3. Which API is called most often
I don't want to integrate some backend software into my code. At least not at this stage.
Hopefully there can be a website for me to upload my logs. After uploading, it would parse my logs and show me some plots. Google doesn't introduce me to such services. Maybe there are some.
At the end, I spent 2 hours parsing the data with Python and analizing them with R. To save myself 2 hours' life, I would pay for such a service.
I thought of making such a product myself. But I want to implement some other idea first.
If you know such kind of service, please let me know :)
Some questions I want to answer from my logs:
1. Which API is slow
2. Which API is broken, ie. returning 4xx or 5xx
3. Which API is called most often
I don't want to integrate some backend software into my code. At least not at this stage.
Hopefully there can be a website for me to upload my logs. After uploading, it would parse my logs and show me some plots. Google doesn't introduce me to such services. Maybe there are some.
At the end, I spent 2 hours parsing the data with Python and analizing them with R. To save myself 2 hours' life, I would pay for such a service.
I thought of making such a product myself. But I want to implement some other idea first.
If you know such kind of service, please let me know :)
A few days ago came across the MIT-licensed GoAccess. It generates both terminal and HTML based reports that should address your needs. You run the tool from the command line, so doesn’t solve the website-to-upload-a-file requirement, but that could be a fun little project (if someone hasn’t already done that).
https://goaccess.io/
https://goaccess.io/
I use GoAccess. Free, open source, run it from the command line and it can spit out an HTML file full of plots. Works in offline or incremental mode.
https://goaccess.io/
https://goaccess.io/
Microsoft’s Kusto database has a free tier and its query language has a “render” command for making plots.
dataexplorer.azure.com
dataexplorer.azure.com
I searched Google for: analyze nginx logs for api stats
The Moesif and Datadog links look promising.
The Moesif and Datadog links look promising.
Thanks. I've spent a few minutes on them while I did my research.
I got the impression that I need to install some software or integrate them into my backend. They are more like system monitor. Maybe I was wrong. That's not how I wanted do it.
I hope to upload my nginx logs and get the results immediately.
I got the impression that I need to install some software or integrate them into my backend. They are more like system monitor. Maybe I was wrong. That's not how I wanted do it.
I hope to upload my nginx logs and get the results immediately.
Perhaps a product that helps indie hackers discover problems worth solving without the hassle of engaging with the real world?
All snark aside, you personally need to experience the problem of you are going to come up with a novel solution people who pay for.
All snark aside, you personally need to experience the problem of you are going to come up with a novel solution people who pay for.
Agree with the second paragraph strongly. I feel like most good ideas come from some amount of background and experience in the problem domain and marrying that with technical ability to build a solution.
That said, because I have the latter ability, I see a lot of examples where people have problems and just don’t imagine a tech solution would help them. This is a blessing and curse for OP.
That said, because I have the latter ability, I see a lot of examples where people have problems and just don’t imagine a tech solution would help them. This is a blessing and curse for OP.
Go visit local businesses. Ask them about what are their pain points. Build for them. Focus on the solution from their POV. You ain’t gonna find shit online.
Many of their pain points could likely be addressed by existing software, cheaper and faster than you could build an MVP for. You could still argue the custom route, however. I read an article about someone who did that and the customers were very happy with the result.
Maybe? The point is to step away from the computer and go to the real world to learn about what problems exist.
Sure, but some people are naturally better at that than others. I'm sure practice helps though.
That’s a very poor argument. No one is born perfect. We all have different levels of potential according to a set of skills. It’s up to us to go put in the practice.
Yes it will be easier for others. I don’t stop coding just because I won’t be to Peter Norvigs level. I still try and try and try and try …
Yes it will be easier for others. I don’t stop coding just because I won’t be to Peter Norvigs level. I still try and try and try and try …
You can practice and get quite good. However, some people are just better at sales, etc, and have already put in a lot of time practicing. However, as you say, all is not lost if you don't know/work with someone like that, it can be done.
People have recommended "scratch your own itch" but I find this is often like trying to see your own eyes.
Imagine a person, or find a person in your life. Create a "persona" around that person's experiences you observe.
So far, if that person is you, great. Now you can see yourself just like you were any other object, with properties and methods, etc.
Then map out the experiences of that persona, and create other personas and do the same for those. As you map out multiple personas, you will start to see shared experiences they have.
As you look through shared experiences you will start to find pain points, and soon you will have pain points that are the same across many personas.
Pain point mapping can be as high-level as "does not have viable future" or as low-level as "cannot remember password" so you want something in the middle like "cannot understand how to get traffic" or a mid-low "forgets to buy milk" or "cannot ..."
This is what people usually partner with a Growth Hacker for, or have a cofounder with ideas already and needs a technical solution to an idea already prepared.
Imagine a person, or find a person in your life. Create a "persona" around that person's experiences you observe.
So far, if that person is you, great. Now you can see yourself just like you were any other object, with properties and methods, etc.
Then map out the experiences of that persona, and create other personas and do the same for those. As you map out multiple personas, you will start to see shared experiences they have.
As you look through shared experiences you will start to find pain points, and soon you will have pain points that are the same across many personas.
Pain point mapping can be as high-level as "does not have viable future" or as low-level as "cannot remember password" so you want something in the middle like "cannot understand how to get traffic" or a mid-low "forgets to buy milk" or "cannot ..."
This is what people usually partner with a Growth Hacker for, or have a cofounder with ideas already and needs a technical solution to an idea already prepared.
If you want the indie hacker life, you have to develop the skills that sustain it.
People have lots of pain points that irritate them enough to pay for a fix, but the vast majority will not frame it in such a clear way or just announce it to the world.
You can take a course like 30x500 (my personal recommendation) that equips you for such a pain discovery and fix definition process. If you're willing to be patient about it, you can also, on your own time, develop your listening skills for that sort of complaint.
Example: from your question, if I could offer a service that lists points of friction or inefficiency that many people complain about and would pay to improve, you would probably pay me to use it and pick a starter for your project. That's a business idea! If I can figure out a reliable, sustainable way to gather this information and then reach out to people in your situation to offer it, I'm on my way to a viable product.
People have lots of pain points that irritate them enough to pay for a fix, but the vast majority will not frame it in such a clear way or just announce it to the world.
You can take a course like 30x500 (my personal recommendation) that equips you for such a pain discovery and fix definition process. If you're willing to be patient about it, you can also, on your own time, develop your listening skills for that sort of complaint.
Example: from your question, if I could offer a service that lists points of friction or inefficiency that many people complain about and would pay to improve, you would probably pay me to use it and pick a starter for your project. That's a business idea! If I can figure out a reliable, sustainable way to gather this information and then reach out to people in your situation to offer it, I'm on my way to a viable product.
Is this thread a paid ad?
For B2C:
1. Is there a community you're already part of? Could be hobby / industry related
2. Do you have an easy way to reach them (e.g. through a sub-reddit)?
3. If people are constantly asking for the same type of information or numbers, is there an opportunity to make a Google Sheets document with useful tabular data? (would only take you a couple of hours)
4. If people keep referencing this spreadsheet you've created, you can throw it into MongoDB and now you've got a) something useful and b) traffic.
Is this simplistic and prescriptive? Yes, but better than create-react-app and hoping to build something people need.
1. Is there a community you're already part of? Could be hobby / industry related
2. Do you have an easy way to reach them (e.g. through a sub-reddit)?
3. If people are constantly asking for the same type of information or numbers, is there an opportunity to make a Google Sheets document with useful tabular data? (would only take you a couple of hours)
4. If people keep referencing this spreadsheet you've created, you can throw it into MongoDB and now you've got a) something useful and b) traffic.
Is this simplistic and prescriptive? Yes, but better than create-react-app and hoping to build something people need.
Love this answer except would recommend Baserow since you want to remove the technical hiccups and get immediate visuals. Combines the Google Sheets and MongoDB aspects.
Airtable is also fine but not self-host capable and getting pretty bloated and clingy with marketing, etc.
https://baserow.io
Airtable is also fine but not self-host capable and getting pretty bloated and clingy with marketing, etc.
https://baserow.io
Love your answer!!! I think "scratch your own itch" is a good answer too, but yours opens up the possibilities not just to my own itches but to other people's itches too. More so, it is sharp and focused. Thank you!!!
You’re asking the wrong question to the wrong crowd. We’re all software engineers, if there was something obvious we’d build it ourselves.
Why not helping to frame the problem better suggesting an improved version of the question?
As an engineer, can't you grasp the spirit of what is being asked and suggest reformulations?
As an engineer, can't you grasp the spirit of what is being asked and suggest reformulations?
I think you're approaching this from the wrong direction. If you have the technical skills and some ideas, you can start proving them out and perhaps pursue one. But there are an awful lot of people with just ideas, most of them bad but not all, who need technical help. I would focus on finding someone how has worked on validation to the point where you're convinced they've got something, then try to join up. The validation is an important key, because it helps prove they're committed and bringing real value to the table. Previously this might have been enough for them to get funding and buy technical expertise, but with changing markets you investing your skills has much more value. If you follow this approach once or a few times you'll learn a lot about discovery and validation and then maybe have something you want to pursue completely independently.
Something I've wanted was a way of setting up any computer I touch to have what I want/expect it to have so I can start working in a dev environment.
I dev on windows, osx, and Ubuntu.
So far I've written a go CLI that is able to bootstrap some stuff for a fresh computer, and tell me what I'm missing.
I dev on windows, osx, and Ubuntu.
So far I've written a go CLI that is able to bootstrap some stuff for a fresh computer, and tell me what I'm missing.
Have you tried Nix[1]? The learning curve was a bit steep for me, but I think I finally started "getting" it and it absolutely solves this problem for me. Now I'm at the point where if I install Nix on any computer, VM, whatever, I can just pull in my configs via home manager[2] and everything Just Works. It's seriously one of the best package managers I've ever used, and I can't imagine going back to anything else. Windows support is only via WSL, so this might be a non-starter for you.
[1]: https://nixos.org/
[2]: https://github.com/nix-community/home-manager
[1]: https://nixos.org/
[2]: https://github.com/nix-community/home-manager
I like the idea, is there anything in this space already?
I'm imagining a GUI where you select from a list of items like "Python Development", ".NET Development" etc, and it somehow pushes out that configuration to your machine(s).
Not sure how many people would be willing to pay for it though?
I'm imagining a GUI where you select from a list of items like "Python Development", ".NET Development" etc, and it somehow pushes out that configuration to your machine(s).
Not sure how many people would be willing to pay for it though?
An IT dept may pay for it to use as provisioning software?
I have an ansible playbook that I can run to add/remove things, set configs etc
My dotfiles has a script called 'update-machine' that pulls the latest playbook, and runs it. I can go from a blank machine to having everything I need in 10 mins or so.
Solutions like gitpod or Github Codespace are going exactly in that direction. Sure, they are not on your machine. But thats exactly why they can make it reproducible, no side effects with other installed software or configurations.
Sounds like you're hungry for Nix.
Is Nix simping the new "I use Arch btw"?
I think its just a suitable tool for this particular usecase.
Fill in the blank on this sentence a few times without thinking too much:
“It’s hard for me to _________________.”
Then, to make sure you’re not BS’ing yourself ask this follow-up question for each answer:
“Have I taken any action to fix this?”
If you haven’t, it’s probably a made up problem.
If you have, there is the starting point for your MVP idea…
“It’s hard for me to _________________.”
Then, to make sure you’re not BS’ing yourself ask this follow-up question for each answer:
“Have I taken any action to fix this?”
If you haven’t, it’s probably a made up problem.
If you have, there is the starting point for your MVP idea…
Don’t bother with decentralization. Nobody knows what that is. You will just end up wasting your time defining a term they don’t to hear because they have already settled on unrealistic expectations that have nothing to do with decentralizing anything.
I have been thinking about how I could monetize performance. I have learned how to load data heavy pages with complex UIs in 0.3 seconds with full state restoration. I have also found a way to test automate that experience with ease and without a giant test automation framework.
Performance is great because the business implications are vast from improved human focus and less fatigue to fewer infrastructure costs and superior user retention.
I have been thinking about how I could monetize performance. I have learned how to load data heavy pages with complex UIs in 0.3 seconds with full state restoration. I have also found a way to test automate that experience with ease and without a giant test automation framework.
Performance is great because the business implications are vast from improved human focus and less fatigue to fewer infrastructure costs and superior user retention.
I'm kind of also searching for that but I don't want to waste efforts in experiments. I want to MVP something that can easily start making money fast.
Not totally sure on how to find it yet, but I'm pretty sure is not just online.
Not totally sure on how to find it yet, but I'm pretty sure is not just online.
Here's a discussion on pain points in the ICU: https://news.ycombinator.com/item?id=33661482
Here.
Fast C/C++ toolchain. Similar speed as tcc but with cpp support. No need to produce super-optimized code, would be nice to have LLVM as an optional backend for release builds.
Fast C/C++ toolchain. Similar speed as tcc but with cpp support. No need to produce super-optimized code, would be nice to have LLVM as an optional backend for release builds.
Scratch your own itch ? :)
Can you tell us what you have tried?
Can you tell us what you have tried?
trying this to build a better Twitter + Feedly + Google Reader + Delicious
https://mutter.cards
https://mutter.cards
Not sure if something like this exists but I'd love a project management software (web or native) aimed at solo developers that prioritizes speed and keyboard shortcuts.
Existing tools cater for teams and organizations, so they come with bells and whistles which I don't need. What I'm looking for is something a tad better than notepad: blazing fast way to create an entry and classify it as feature, bug, research etc, add some tags like frontend, backend, UI, etc and add some notes. Obviously there will be a little more like a great search, a way to quickly switch between projects etc but the focus should be on simplicity and speed of use. Ideally pretty much everything should be doable via keyboard. It's ok if it's opinionated and has limited functionality. But as long as core task management features are seamless, I'd be happy.
I'm working on a start-up idea right now and thought the above would be really helpful (own itch). Also low key thinking of building it myself.
Existing tools cater for teams and organizations, so they come with bells and whistles which I don't need. What I'm looking for is something a tad better than notepad: blazing fast way to create an entry and classify it as feature, bug, research etc, add some tags like frontend, backend, UI, etc and add some notes. Obviously there will be a little more like a great search, a way to quickly switch between projects etc but the focus should be on simplicity and speed of use. Ideally pretty much everything should be doable via keyboard. It's ok if it's opinionated and has limited functionality. But as long as core task management features are seamless, I'd be happy.
I'm working on a start-up idea right now and thought the above would be really helpful (own itch). Also low key thinking of building it myself.
My partner did a DNA test through Ancestry.com because she wanted to know her ethnicity. She was the subject of a closed adoption. When she got her results back, she got a list of people who opted-in to sharing DNA match status with others. Her matches were good enough that it was very possible to find her biological parents.
There are a lot of techniques to building familial profiles from this list. One is called “Leeds” where you place individuals into groups who they themselves match with. You can then color-code it and get a visual distinction of families. There are other techniques, too: grouping matches by ethnicity, and by ancestor.
Obviously, I didn’t want to do this by hand like everyone else. I sought out existing software but the two solutions required me to login to Ancestry (and not via oauth), and run a client on my computer like it’s 1999.
No thanks.
I wrote some Ruby scripts, it didn’t take long. I joined some Facebook groups which where helpful. I tried to be helpful with what I had learned, so I explained the techniques to others who had posted in the group.
I noticed that not everyone does this on a computer with two monitors, like I have. A lot of people don’t even own a computer these days, contrary to what HN would lead me to believe.
One person was especially... I don’t want to say inept, but desperate. I offered to run my code against their matches after they added my account as a collaborator to their test, allowing me to see their matches without needing to give me their login. I did, they said thank you.
The next day I woke up with 42 messages in my Facebook inbox asking if I could do theirs.
I spent the weekend putting together a shitty website that automated this and sent the link. Everyone was very grateful. I listened to them and built it out a little more. Some seasoned “search angels” (people with a knack for this kind of sleuthing) asked me to build out some other reports. I created them, and asked these people to record some of their workflows so I could see how I could improve their manual processes. A Chrome extension was born to inject some code into Ancestry for better UX.
Little did I know that when I took a DNA test I would learn that my dad isn’t my father. My mom passed away three years ago, and despite us being best friends, she never told me this fact. I have since learned that she mentioned this fact to others, but never disclosed a name (and nobody ever pressed).
I myself am on this wild goose chase now, and my new search angel friends are on it with me, though for the last few months we have been collectively stumped and brickwalled.
What I’m trying to say is that there are problems all around us. We just need to listen for them, and question why things are done the way they are done. If there’s any sort of wiggle room, or gray area as to why, it might be a problem worth solving. It might not be a sexy technical problem worthy of the front page of HN; it might not even have enough bros in the community to boost you on ProductHunt. The case for using a new hot trendy stack might be lost. It might not even be all that profitable.
But if you adjust your expectations, and find something that genuinely interests you instead of someone else, you should get enough satisfaction to make it worthwhile.
My matches, in case anyone is curious: https://sherlockdna.com/ancestry/leeds_reports/99a62d1cb68b7... (sorry mobile users, I haven’t quite figured out the UI here)
There are a lot of techniques to building familial profiles from this list. One is called “Leeds” where you place individuals into groups who they themselves match with. You can then color-code it and get a visual distinction of families. There are other techniques, too: grouping matches by ethnicity, and by ancestor.
Obviously, I didn’t want to do this by hand like everyone else. I sought out existing software but the two solutions required me to login to Ancestry (and not via oauth), and run a client on my computer like it’s 1999.
No thanks.
I wrote some Ruby scripts, it didn’t take long. I joined some Facebook groups which where helpful. I tried to be helpful with what I had learned, so I explained the techniques to others who had posted in the group.
I noticed that not everyone does this on a computer with two monitors, like I have. A lot of people don’t even own a computer these days, contrary to what HN would lead me to believe.
One person was especially... I don’t want to say inept, but desperate. I offered to run my code against their matches after they added my account as a collaborator to their test, allowing me to see their matches without needing to give me their login. I did, they said thank you.
The next day I woke up with 42 messages in my Facebook inbox asking if I could do theirs.
I spent the weekend putting together a shitty website that automated this and sent the link. Everyone was very grateful. I listened to them and built it out a little more. Some seasoned “search angels” (people with a knack for this kind of sleuthing) asked me to build out some other reports. I created them, and asked these people to record some of their workflows so I could see how I could improve their manual processes. A Chrome extension was born to inject some code into Ancestry for better UX.
Little did I know that when I took a DNA test I would learn that my dad isn’t my father. My mom passed away three years ago, and despite us being best friends, she never told me this fact. I have since learned that she mentioned this fact to others, but never disclosed a name (and nobody ever pressed).
I myself am on this wild goose chase now, and my new search angel friends are on it with me, though for the last few months we have been collectively stumped and brickwalled.
What I’m trying to say is that there are problems all around us. We just need to listen for them, and question why things are done the way they are done. If there’s any sort of wiggle room, or gray area as to why, it might be a problem worth solving. It might not be a sexy technical problem worthy of the front page of HN; it might not even have enough bros in the community to boost you on ProductHunt. The case for using a new hot trendy stack might be lost. It might not even be all that profitable.
But if you adjust your expectations, and find something that genuinely interests you instead of someone else, you should get enough satisfaction to make it worthwhile.
My matches, in case anyone is curious: https://sherlockdna.com/ancestry/leeds_reports/99a62d1cb68b7... (sorry mobile users, I haven’t quite figured out the UI here)