Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: What's a side project you built to make money that hasn't?
446 points by notoriousarun 11 months ago | hide | past | favorite | 582 comments
A friend pointed out a bunch of the 'tell us about your successful side project' threads suffer from a survivorship bias. They're still great for inspiration, but I suspect we could learn a lot about challenges and wrong approaches from each others' failures. So what's a side project you built hoping to generate revenue from it, that hasn't actually earned you much / any money?

Why do you think it hasn't been as successful as you thought it would be / what would you do differently if you did it again? How much time/money did you spend building it, and what kind of iterations / improvements did you make to try and salvage it?

Appreciate any and all answers!

In 2004 my brother and I developed a buttplug shaped like George W. Bush - we called it the Bushplug. Manufacturing was more expensive than we had expected but we sold about 100 of them - presumably as novelty gifts. Our price point was too high for a novelty gift and the nose was a little too pointy to be an enjoyable sex toy. We made our money back and had some fun with being featured on BoingBoing and Fleshbot. Someone tried to sue us. We sent the last of the Bushplugs to the Smithsonian's presidential museum. We made our investment back but didn't become buttplug millionaires.

The other alternative to the pointy nose theory was that there just weren't a lot of people who were into having W up their butt, but we felt that must not be the case.

> The other alternative to the pointy nose theory

you'd think it would have been much more successful with Obama


Obama’s nose is a bit flatter than GWB’s, so would poke out a bit less.

Someone did this again in 2012, making 3D models based upon each of the Republican candidate polling numbers: https://www.huffpost.com/entry/gop-butt-plugs-are-gallup-dat...

As the other HN Buttplug Poster who is also having problems making money with their product, I salute you.

(And yeah this is why I stay in software, manufacturing even on non-tech toys is still $$$$)

What niche does your butt plug serve?

We weren't sure if our main demographic would be 'straight' Republican congressmen or angry gay Democrats - this may have been an issue.

Thank you for this.

Lol! Those look a lot like ours did but we didn't have color - it really makes the difference.

Did you ever think of making a Dick Cheney Dildo™ to complement the set? :)

If we had that, we 'wood' have certainly succeeded! Hey...that makes me realize the Cheney Dick could be made out of wood so the splinters would stay with you forever.

This is hilarious. Thank you for the laughs.

You should look up the pictures of it - there must still be some out there...

Almost uncontrollable LOL here

Right idea, wrong time. My friend owns a sex shop in Provincetown, MA and has sold tens of thousands of Trump-plugs over the past 4 years.

location, location, location


I built a Tinder-like matchmaking app where users log in with their Facebook account and the app then auto-matches users by their FB likes. I thought taking the hassle out of swiping could give my app an edge over Tinder.

It was working fine for a week or so before Facebook caught up and revoked my API keys (effectively killing the app). They didn't give any reason besides some vague recommendation to review their Terms of Service. A few months later at their F8 developer conference they announced they'll be launching their own matchmaking service which will work pretty the same way I built my app.

Moral of the story: never trust big tech.

P.S. I open sourced the app a few years ago: https://bitbucket.org/stonepillarstudios/workspace/projects/... Feel free to fork & revive.

I've heard that it's a good rule of thumb to not develop something that is just an extra feature of an existing behemoth (e.g. Atlassian plugins etc., but also the thing you built). If you're failure, then you've wasted time and money and if you're a success, the behemoth will just reimplement your work as a part of their platform.

Pretty much the only way to win it so make something that provides marginal revenue, which makes reimplementation not worth it for the giant (but potentially worth it for you). I suspect a lot of Unity and Unreal plugins and authoring tools reside in that niche.

Wordpress/Automattic bought up successful plugins like WooCommerce and just integrated them.

Automattic is not in the same league as FAANG, it doesn't really apply.

Marginal revenue is relative though. For the developer, if it makes $500k over the product's lifetime, it can be worthwhile. For Facebook, if it only makes $500k, it's a waste of resources.

Couldn't one expect at least a job offer?

Maybe, but working hard for months/years a chance at a job offer doesn't sound that amazing anyway.

BTW I worked in a startup once and a guy, who was developing opensource plugin for our product, applied for a job with us. The owner (a known blogger BTW) said "Why hire him if he's already working for us?"

I think that was the owner being shortsighted. If the plugin was something that a reasonable fraction of your users would like to have, then very shortsighted.

Not only you had a person that already knew much more about your product than the average potential hire, but a person that probably was interested in your product (and maybe knew what users wanted) too.

"Why hire him if he's already working for us?"

Because you do not want him to abandon the project if it is useful to you?

"I built a Tinder-like matchmaking app where users log in with their Facebook account and the app then auto-matches users by their FB likes"

This is what most dating apps do, except Tinder (trying to do a clustering in an n-dimensional space). That Tinder outclasses most other dating apps is proof that your approach wont work. We don't look for people with similar interests for dating. Unfortunately we don't chose who or what we like in a partner.

This like the case of Amazon antitrust

current Tech giants are need to be replaced

> current Tech giants are need to be replaced

There's no reason their replacement will be any more moral. They should be dismantled imho.

Should the tech giants really be punished for being too successful? I don't think so. They got to their size by being better. Nobody thought yahoo would ever be unseated. Yawho?

Antitrust laws have been around for a long time. Competition is essential for a free market to function optimally. You can disagree, but that's why they exist. Its not to punish anyone.

Wow that really sucks! Sorry to hear that. I suppose you would've needed patents on the idea to stop them from stealing it.

Or they were already working on it and hadn't announced, and he would have spent all the time and more money working on the patent only to see Facebook launch the same thing before he even filed it. Assuming there wasn't already a patent out there that covered this.

If you had patent, you could go after FB and have some out of court settlement.

In reality you can't. The upfront cost of defending a patent is huge, like hundreds of thousands.

Patents are fundamentally broken and really don't work like most think or want them to.

Sounds like you should be taking this case to the Congress.

In 2006/2007 I built HTMLButcher, a C++ desktop application to slice PSD/image website designs made by designers to HTML. In these years sites where made with tables, so for slicing a design I had to cut the images and fit then into borderless HTML tables. I took 2 years to build it, and in this timeframe people started building sematic websites with CSS, and abandoning table-based designs. I managed to sell 100 copies, for some reason 90% to India. After some years without selling nothing, I open sourced it: https://github.com/RangelReale/htmlbutcher. The good thing is that I REALLY learned C++ with this project, and this knowledge was the basis for my current company where I made a digitalsignage application in C++ that runs in some of the largest DOOH places in Brazil, in Windows, Linux and MAC (targets of HTMLButcher). So in the end it was a good investment!

A zillion years ago I had a housemate who was a brilliant programmer and could smoke himself into the Zone like nobody's business, and one of his weird little side ventures was a Mac app to cut up Illustrator files and lay them out in HTML using tables, because that was the state of the art.

As best I can recall, he wrote it in CodeWarrior over a couple weeks together with another guy who presumably knew Illustrator, and Adobe bought it before they had to deal with anything like a business plan.

Was it possibly integrated into Adobe ImageReady [0]? It had a feature like what you're describing where after you finished designing your website, you could define cutouts and export as images+HTML.

This is actually how I first got into programming when I was younger. I made my first site in ImageReady, but it couldn't do all the other cool stuff other sites did like comment sections and logging in, so I went on a quest to learn how to do all that and the rest is history :)

[0] https://en.m.wikipedia.org/wiki/Adobe_ImageReady

Edit: actually, I found an example from 13 years ago https://youtu.be/qCq0JiwMcJs

> he wrote it in CodeWarrior

Wow, that's a name I haven't heard in 20+ years. We used CodeWarrior Pro for my Pascal class at a local community college that I took during high school, because my high school was so small, I literally ran out of classes to take by senior year, so I had to take a programming course at a CC that was 10 miles away. Had to get a waiver to leave school before noon to go to the class because state policy was that students can't be released from school before noon, and because all my classes (all 3 of them) were over by 11:30.

> could smoke himself into the Zone like nobody's business

Can you elaborate on this?

Smoking weed to get into the proverbial, "flow" state.

What is the Ballmer Peak[0] equivalent for weed?

[0] https://xkcd.com/323/

The Snoop Summit

Also take a look at the marvelous table-based design of the app website: http://rangelreale.github.io/htmlbutcher/index.html

I would have paid a lot of money for that back in the pre-CSS days, assuming it worked well. Was it purely image-based or could it recognize when an area could be rendered without using an image (like a single color zone/etc.)?

You could select a kind for each cell, "None" would not generate an image, and you could select a cell color or html tags. It could also be "Image" or "Mask", with Mask being an inner table within a cell.

You can see this on the main screenshot on the website: http://rangelreale.github.io/htmlbutcher/index.html

That’s cool. Everything about the design fills me with nostalgia and takes me to the good old days of the Micro ISV, the Joel on Software forums, and my own, now unsuccessful, plans for world domination.

I'm a little confused. That feature has been built into Photoshop since at least CS1 (2003), maybe earlier.


That's where you start. But to make it responsive it takes more work. If I remember..

For a 9x9 basic table grid you would set the width 100% of the top and bottom image do the opposite for the sides and fix the corners. All of these tricks involved choosing designs that could work.

It would have been a tool I would have liked.

Also webdesigners loved to change the layout mid-project and it was a big hassle to migrate the slicing from one PSD to another.

Not counting that some websites had dozens of PSD files, it was a pain.

It never worked well. The html was full of issues and the kind of image slicing was for backwards compatibility with email clients. However PS did have a way to write scripts with JS and extend or even change the HTML generation.

Photoshop had a kind of high price point just for cutting up images.

Can you share the name of the Digitalsignage company? I am in market for such a solution

I made a plugin for photoshop around 2010 to do the exact same thing. It has a slicer tool, and an API for it, so spitting out HTML was easy, but just used it as an internal tool for one of my first web dev jobs.

I would have used this in ~2005. Seems like you were just slightly late to market. CS2's built in tool to slice PSD -> HTML kind of sucked and required a lot of post-processing work.

Apart from the table based design, did you feel any competition from services like CSSilize? Was it difficult add support for the DIV based design?

I added and AP tablebless (absolute positioning) option, but this was only simulating tables with CSS. People stopped using large and sliced images with CSS, so it wasn't worth it.

It wasn't a big serious project or anything, but for my first income generating project ever I made a small single player word game for Android that I should have tried to monetise more. Android had only been out for a couple of years at the time so it was easier to get noticed then.

The game was free with an ad-banner during gameplay. I was making a decent amount in ad clicks and also had players emailing me for a paid version that just disabled the ads. Instead of jumping on this to release a paid version and expand the game with more features, I ran into perfectionism issues + decision paralysis releasing any changes.

The game blew up for a few weeks and I got something like 0.5 million downloads eventually but didn't do anything to monetise it more so felt like I missed a big chance.

It was a great feeling checking analytics though to see that there was literally years of collective gameplay time being logged, and I've had several emails and reviews from players saying they've been addicted to playing the game for years.

My biggest lessons are probably:

1) Doing/releasing something is always better than doing nothing so don't let decision paralysis get in the way. "Don't let perfect be the enemy of good".

2) Don't fear releasing or making changes because you might get bad reviews. You'll never please everyone and even the perfect app will inevitable get some brutally unfair and weird/crazy reviews.

3) Some people are obsessed with word games!

Anyway, as a lockdown project for fun, I started making a new web version for mobile + desktop. :)


I just spent way longer playing this than I meant to. I know it's similar to a lot of other word games, but it's really fun and satisfying, and is meaningfully different than other word games I've played. Good luck with it!

Thanks! Yep, I found it interesting how slight tweaks to rules for word games gives them a really different feel and incentivises different strategies.

I need to add global high score tables. The highest score I know so far is close to 3000 which is nuts.

Wow, what a fun game! Projects made for the fun of it tend to have a certain charm to them compared to projects made for profit.

Thanks! Yeah, that's part of what derailed me trying to make a paid upgrade for the old free version. I had to think of what features would be locked away in the paid version, how the player would be nudged to upgrade etc. It's a fair amount of coding work and planning, and it's not fun compared to working on the actual game.

FWIW, personally, I would pay for no ads and maybe brief definitions of the words that I have just crossed out, but I would be turned off by microtransactions like boosts, extra life's or skins.

Cool, thanks! Yep, I hate soulless plays like "buy gems to play more", "wait until tomorrow to get more gems to play", "invite your friends to get more gems" etc.

Word definitions would be awesome. Are there any good free sources for this? That also have offensive words flagged?

I'm not sure how to do a free version of the game that isn't so good that nobody upgrades though to be honest. With the competition on mobile now as well, I feel freeium is a must to get tractor too.

en.wiktionary.org might be good for this. It’s free and marks some words as vulgar.

It's a really fun game. I was a little frustrated by some words which didn't seem to be recognized - I think 'rome' and 'zen' were two, whereas much more obscure words were accepted.

Thanks! Proper nouns (generally, words that start with a capital letter) are disallowed so Rome won't work.

Hmm, Zen was recently added to Scrabble in 2018 though as it's a proper noun but has started being used as an adjective:


Your understanding of the meaning of proper noun is inverted :)

Doh, fixed (hopefully). :)

Neat web version, but yikes did that thing light up my laptop's fans and make FF use nearly 100% of two 2.8ghz i7 cores here. Seems a bit ridiculous considering what's going on.

Argh, I'll look into it because it shouldn't do that. It's using DOM + simple hardware accelerated animations intentionally (over something like canvas) and no polling to keep CPU usage down. Runs fine on Chrome on Mac.

Yeah well, in unaccelerated scenarios like mine that translates into software rendering on the CPU, which there seems to be a whole heck of a lot of happening despite nothing really changing on-screen.

Thanks, I'll need to check some more because the Chrome performance tab is saying 95% idle when nothing is being selected as expected for me.

Either way, I'd be keen to add a battery saver mode because it's not the kind of game I'd want to devour battery on a mobile.

Looks good. On my tablet PC, it's not letting me play by touch. For users with both mice and touchscreens, one has to handle the touch events that browsers have.

Thanks! Hmm, which OS and browser? When I test with the "laptop with touch" dev tools device on Chrome + Mac it works as expected so not sure how to replicate.

Chrome latest on Windows 10. The key point/problem is that because I have both mouse and non-faked touch, the touch events aren't necessarily converted into mouse events, as happens in many other circumstances. Though a lot of this is conjecture.

Fun game, even with trackpad I can do diagonal words. Nicely done.

Thanks, it's tricky to get the selection of diagonals working accurately like that. If you inspect the HTML, you can see each tile has an invisible 45 degree rotated clone on top of it that's used to detect when a tile is selected instead of the boundaries of the original tile.

Cool game. Spelling June didn't register a valid answer.

Thanks! Proper nouns (generally capitalised words) are disallowed so June won't. Not sure how best to communicate this in the game though.

The original had a wall-of-text pop-up of the rules when you first started that I want to avoid and I'd rather players picked up the rules organically. Maybe showing a "proper nouns and capitalised words not allowed" message when you spelled one would be a nice approach but a list of such words would be massive and inaccurate.

I think people generally figure out this rule as they play more but I admit it's not ideal!

It is interesting because the same happened to me while trying the game.

If something happens to almost everybody, why don't you change your game?

When I design a complex system like a game or computer programs, once the system starts working on their own, it talks to me. I learn a lot from the interaction with it.

This is specially important when you have millions of users. They communicate to you what the next step is much better than the best design you could draft isolated on your own.

Yep, I'm up for making change based on player behaviour and ideally the game would have zero instructions. Proper nouns come up often in feedback too.

Adding proper nouns has a few issues though:

- Where would you get an accurate dictionary of proper nouns? This would need to include e.g. people names, places, brand names. If it wasn't constantly kept up to date it could lead to more frustrations than just disallowing them.

- There's so many proper nouns it would bloat the game download. The English dictionary used right now takes about 3MB, which is the limiting factoring getting the game download lower than 1MB zipped. I'm failing to find an estimate but if you include proper nouns, the dictionary would be an order or magnitude bigger. Maybe less of an issue for mobile app store downloads but I want the game to load fast when shared online. A web service to check the words won't work for offline play either.

It's feeling like a "dictionary words only" message somewhere is the only practical solution. I could adapt the (single!) tutorial message at the start of the game to be "swipe in any direction to spell dictionary words" but "dictionary words" would probably be overlooked. There's no menu screen at the start because I know people are impatient. :)

Regarding size, I like to remind myself how much bandwidth it takes to watch just seconds worth of Netflix or YouTube.

Not everyone has a fast connection though and you'll try the patience of new players by making them wait just a few seconds. I also like optimising so if I can make the transfer size x10 smaller I will. :)

Recent startup failure was ZguideZ - an app that allowed locals to create their own digital tours, charge what they want, and then collect the majority of the profit on tours sold.

It failed because it was a far bigger project than I was able to manage as a solo founder - though I tried. It was beyond my ability to code it and I didn't have funding to recruit quality developers. I hired a budget team and we cut corners to stay in budget. I was unsuccessful in recruiting a team - partly because I had taken on a cofounder who decided he would rather surf than work on the business development end of our product - without a founder agreement, he was dead weight that scared investors and potential team mates away. In desperation for help, I began working with a veteran who had recently separated from Army intelligence - he had undisclosed mental issues and when our business plan was made a finalist in a university bizplan competition, he accused me of being a spy sent to retrieve classified information from him and sabotaged our meetings with investors, potential hires, and the competition administrators. He withdrew us from the competition and sent out insane accusatory emails to our bankers and advisors.

I attempted to carry on but the shutdowns of tourism in Hawaii and sheer exhaustion over my co-founder mistakes led to shuttering this project we thought would be the next AirBnB.

All told, this was a budget MBA program for me which ended up costing about 1/4 what a quality MBA would have cost me and probably taught me far more.

Lessons learned were: 1) the importance of a founder agreement 2) the importance of doing enough due diligence to understand the true scope of a project and then doubling or tripling the amount of work it will take to achieve that scope 3) the importance of working with the right people and refusing to settle when it comes to product or team

> I had taken on a cofounder who decided he would rather surf than work on the business development end of our product - without a founder agreement, he was dead weight that scared investors and potential team mates away

This is, unfortunately, very common in startup communities. There are a lot of people who attach themselves to startups, discover that they're not interested in putting in the work required, and then decide to coast as long as possible when they realize it's not so easy for their partners to cut them out.

These people are often very charismatic and are experts at inflating their previous credentials. Always do a deep dive on potential cofounders' backgrounds, including backchannel references. Always establish cofounder agreements and, most importantly, at least a 1-year vesting cliff on any equity, no matter how much you like the other person. If they aren't pulling their weight at the 9-12 month mark, do everything to remove them from the company before they have any vested equity.

> In desperation for help, I began working with a veteran who had recently separated from Army intelligence - he had undisclosed mental issues and when our business plan was made a finalist in a university bizplan competition, he accused me of being a spy sent to retrieve classified information from him and sabotaged our meetings with investors, potential hires, and the competition administrators. He withdrew us from the competition and sent out insane accusatory emails to our bankers and advisors.

Mental illness is also, sadly, disproportionately represented in startup communities. I think it's a combination of factors: People who suffer from certain untreated mental illnesses like you describe are more likely to be unable to maintain a normal career, so they seek out alternative career paths like startups where they don't technically have to report to a manager. Certain mental illnesses can also come with bouts of grandiose thoughts, delusions of grandeur, a penchant for risk-taking, and other traits that can easily be confused for confident, ambitious startup personalities.

This is a perfect example of the importance of operating agreements and vesting cliffs, as well as due diligence on anyone you might go into business with.

AirBnB has something like that called "Experiences." I'm not saying that to be critical; if anything it should be affirmation that you were onto something, i.e. there really is a market for such things.


During the pandemic, AirBnB has also been helping people organize online experiences.

Definitely a market for it. Another Hawaii company, Shaka Guides has a similar concept though with voice-actor recorded tours and they seem to be doing well. Our concept was to have locals sharing their history and knowledge on a self-guiding tour.

"Experiences" isn't the same as this idea at all.

AirBNB Experiences is things like "Come swimming with dolphins" or "come on a wine tour".

This idea seems more like the voice-guided tours you can get in museums etc but creating a market for them. It's probably a decent idea, but possibly complementary rather than competitive with what AirBNB is doing.

Amazon has recently launched a virtual experienced platform and it’s actually quite fun. (when not hamstrung by technical difficulties)


This idea seems really similar to Detour, created by Andrew Mason and later acquired by Bose: https://techcrunch.com/2018/04/24/bose-acquires-andrew-mason...

The idea of running it through a pair of AR sunglasses is wild! We were doing it strictly on smartphones - some of the biggest headaches came as a result of Apple's App Store process - we had to rework everything and since the back end was independent that meant we had to rework the android version as well.

I really liked Detour. Had some interesting walking tours of SF thanks to this app.

> due diligence on anyone you might go into business with. > importance of operating agreements and vesting cliffs

I'd argue it's more important to work with someone you've worked with in the past and know you can trust and will work hard. The operating agreements and due diligence are there for the worst case scenario, but from your story it seems pretty clear they were questionable co-founders to start with.

I think this was a great idea btw, sorry that it didn't work out. Before covid I always used to try and find audio walking tours for cities, and alternative ones to the normally rather dry official ones for museums, but there really isn't much out there.

Wow, sounds like a rollercoaster. You sure got some stories out of that one though!

I like the concept a lot, shame you never got to find out if it would have worked.

Really powerful experience that I am grateful for (though I wish my hair would grow back - lol). At the moment working on a much more manageable project with a great team (Iwahai.com) and maybe in the future there will be an opportunity to revisit the ZguideZ concept. I still love the idea, but managing a worldwide network of independent guides is more than I want to take on at the moment.

I built a really simple way to poll 50 random people in the US.

It's called This or That and functionality is currently very simple. You submit a question along with two images via SMS and you get your answer back usually within the hour.

I've had a few people use to test new logo ideas, to ask which of two TV shows to watch, or which outfit looks best. So far no one has paid for it, only a few hundred free users, but I think there's something here I just haven't marketed it to the right audience yet :)

My goal is to make this usable via Slack next year and let teams use it to trial new marketing campaigns or other run other small tests before launching. https://www.thisorthat.ai

I like this idea! And not only because I have a long-running art project called This Like That.

But I'm not sure about using Mechanical Turk. If the goal is to have unbiased opinions, is that a good place to source them?

I would assume MT participants are trying to complete tasks as fast as possible. If I'm on MT and you ask me "A or B?" my incentive is to give you an immediate answer. I have no incentive to align that answer with my actual opinion.

Also I wouldn't expect MT to give me a particularly random sample. I mean, if I'm wondering how my B2B SAAS logo is perceived, is an answer from MT going to overlap much with the opinions of my target audience?

It seems like a fun thing to try as a toy, like rolling dice only with humans. I'm just not sure I'd use it for any question I really cared about.

Make it useable by asking people to answer questions in exchange for getting to ask their own question?

Reaching critical mass may be harder, but I think it would be more random than mechanical turk.

Agreed, answering in return for ask credits is a good idea. Maybe 10 answers = 1 question or something like that. If you make the ratio too large then people will just select anything to get their credits. Another option is to only award a credit if their answer was in the majority.

You can also charge people for the ability to ask a follow up question like why they made the decision they did.

Heh, I built something similar years ago. Even got a YC interview. It was called decisioncandy. Originally targeted logos but switched to apparel...

I think the Slack idea is brilliant though, and I could totally see that taking off for a wide variety of things.

Super interesting. Any other ideas from your testing, or feedback from the YC interview on where there could be traction here? If you're willing to share, of course :)

Hmm certainly willing to, but I don't have much. Mostly just that professional graphic designers didn't seem like the right audience since they already have networks of peers who can give better feedback. Oh, and the apparel-industry-specific stuff from below.

But something like an engineer asking a colleague which web design they just hacked up looks better would be a much better fit.

I had an idea once about something similar, I called it Design Picker, where you'd basically use some combination of crowdsourcing and "algo" to help you make design decisions. I had the domain for years, but I could never come up with a version of the concept I thought actual designers would actually use, so I never built it.

Did your YC feedback yield hints as to why you were rejected (I am presuming)?

Yes, a very clear answer! The problem we were solving was that apparel firms have really long lead times: they design their lines for a season 9-12mo before the customer buys it. That means they can't iterate on their designs; DecisionCandy would let them do so in a lightweight way, by showing their sketches/previews of pieces to their customers, and doing a "which do you like more" survey to score the items.

The problem here is that it takes 9-12mo to find out whether the app does any good, and YC is a 3mo program! We were only just about to sign a pilot deal at the time, so the timing wouldn't have made sense.

And indeed, that long of a feedback loop is realistically too slow for a VC startup. We closed up and I "pivoted" to something totally different a few months later.

Hm.. I worked on a piece of software with similar needs of polling random people for their opinion. Have you considered using Amazon Mechanical Turk?

I imagine some bias would be there, but probably manageable. In return you'd get the ability to poll way more than 50 people for a very reasonable price.

It says on their site they are using Mechanical Turk:

> To do this, we take your question and use Amazon's Mechanical Turk platform to survey 50 random people.

Thanks! This is why I keep saying that all my good ideas are already taken! ;)

I participate in a poll/survey portal as a reviewer. Just to give you an idea of how it works: we the participants are free to register and get some points after every completed poll/survey (the amount of points depend on how long it took). With these points we can access to a "shop" of small prizes, like headphones, small cameras, books, etc. I assume that the companies making the polls are the ones paying and probably they select the demography of the people to survey (we're asked about this info on the polls). Maybe a model like that can work for you too.

Love this idea but are the surveys randomized or quality controlled at all? I have tried 4 surveys, all different and the results seem weirdly consistent. Option A always seems to win out with ~60% of the vote (60%, 64%, 60%, 57%). I even reversed the two images in one test and whatever was in position A won.

Thanks for flagging this! I had quality tested a month ago (nothing scientific) and tried multiple different image variants, placement (A vs. B) and they did remain consistent. That said, I refunded your orders while I take a look at quality control. Thanks again for trying out and letting me know.

Perhaps entrepreneurs wanting feedback on ideas could be good niche to start with?

If there's anything I learnt from discussing my ideas with family, friends and others, it's that finding the right people to bounce your ideas off of is crucial.

I think Will Wright (famous game designer) described this phenomenon well. To paraphrase, every time he would try to describe a new game to someone else, that person would create a game design in their head and play their game in their head, and then tell you what their thinking is. However, what you want, is someone to play _your design_ in their head. That's very hard to do without a prototype. For example, basically everyone told him that idea for The Sims is terrible. Only when they played his game did they see the light.

Interesting. How do you incentivise the opinion givers to actually consider the options instead of just efficiently (randomly) selecting an answer?

How about a reward for choosing the "correct" answer.

If the side with the most votes is the side you chose, you get 2 points. If the side with the most votes is not the side that you chose, then you get only a single point.

Have you tried removing the free trial?

NN-512 (https://NN-512.com) generates stand-alone C code for AVX-512 neural nets

The idea was that user companies (or even the hardware companies: Intel, AMD) would pay me as a contractor to implement deconvolutions, transformers, etc., as extensions to the free/open core I released (which includes very sophisticated support for all forms of dense convolution, fully-connected layers, pooling, batch norm, activations, etc.)

The free/open core supplies all the basic operators, but many users would need extensions of the core to support their particular networks (or even just software to feed their parameters into the system and get results out) and hardware companies benefit from the software's existence because they sell chips and NN-512 makes the new vector units on their chips more useful

But it didn't work out that way. NN-512 did generate offers to work on other projects (a generous offer from Intel, for example) but nothing that extends NN-512

Why? There are numerous free end-to-end deep learning tools developed by large, company-sponsored teams (Nvidia, Facebook, Google, Intel, etc.) so a specialized tool that requires integration effort is unattractive. That's my guess. So beware, if you're working along similar lines

I saw this in one of your other posts and it seems super interesting to me.

1. Improve the web site slightly (add a contact method and pricing info) and make it clear that it's a company offering an enterprise software tool. Programmers will be viewing the site, so it doesn't need much more than is there already.

2. Charge a flat $7,500/year for a commercial usage license and increase it as the functionality increases. Offer email support and updates. Many programmers can buy a $7,500/year piece of software without their management even asking a follow up question.

3. Write some blog posts showing how to use it and why it's good. Some benchmarks (LOC, performance, etc) vs alternatives would be a good source of posts. Real world examples that people can copy/paste to start with is probably the most useful.

4. Submit the blog posts to HN and some subreddits.

5. Profit (maybe)

> Many programmers can buy a $7,500/year piece of software without their management even asking a follow up question.

I’d like to know where you work.

No place I’ve ever worked would approve that without a lot of questions.

> He is not wrong

Thank you for this thoughtful reply!

I work in the space and was impressed with NN-512 as there's a painful gap in inference cost between CPU and GPU that doesn't have to exist. Intel and AMD are really missing a boat here, most other companies have enough cash they just go to GPUs, academics rarely sling low level code even in CUDA let alone AVX-512, and other than Fabrice Bellard's work few I've seen few go that low level.

My suggestion would be to focus on an initial use case where a very limited low cost / high efficiency CPU model can provide massive advantage. NN-512 should be the framework that expands from that Redis like core. The limited use case tactic is what I'm focusing on[1], mainly as I have a particular application and have less technical brilliance than yourself so need to focus ;)

An aged but still relevant example is the early word2vec work which was (and still is) frequently better to throw onto CPUs than GPUs. A well tuned implementation is not only advantageous on CPU but can win out in many scenarios where cost / latency / ... are important.

Congrats on the project though! I'd be curious for your thoughts for the future if you ever want to chat =]

[1]: Initial experiments written up as a tutorial with Rust and ISPC for a specific CPU based NN task - https://state.smerity.com/smerity/state/01E8RNH7HRRJT2A63NSX...

In your [1], are your input/output arrays aligned the same for the fastest SIMD and ISPC runs?

Not in that codebase as it was a tutorial / wanted to ensure it's callable from safe Rust code so stuck with `_mm256_loadu_ps`. That code was just playing with dot product like lookup over vectors on CPU. The code I'm more interested in is trying to cram models into ~L2/L3 cache such that a CPU optimized model can be trained on GPU to be deployed on CPU.

How about a tool that generates C (or Java or X) code from TensorFlow or PyTorch models? That way you can train a model offline, and then export it to your preferred language for serving/production. For bonus points, figure out how to get the generated code to use the GPU properly.

Addressing your last sentence, I am actually a GPU programmer by trade. GPUs are my specialty, and much of my professional work has been in Nvidia's CUDA

But AVX-512 is a real step forward, if you look at the price/performance ratio for real-world workloads. So AVX-512 is the focus of NN-512

GPU cloud compute is almost unbelievably expensive. Even Linode charges $1000 per month, or $1.50 per hour (look at the GPU plans: https://www.linode.com/pricing/#row--compute). It's really hard to keep that GPU saturated, which is what you need to do to get your money's worth

An AVX-512 Skylake-X cloud compute instance costs $10 per CPU-core per month, or 2 cents per hour, at Vultr (https://www.vultr.com/products/cloud-compute/). It's easy to saturate that core

Think about it this way: a GPU is really just a cluster of SIMD units with fast memory. A GPU thread variable is just like a SIMD lane, and a GPU warp variable is just like a SIMD vector register

Nvidia's SIMD pseudo-instructions are called PTX and they are like a 1024-bit version of AVX-512. An AVX-512 core is like a general purpose CPU with a 512-bit GPU core built in

So paying for a single AVX-512 core is like paying for part of a GPU (16 threads), plus the general purpose compute you need to keep that GPU part supplied with work

I've been interested in your project. I don't think you're realizing the potential of the parent's suggestion.

You motivate the work through bottom-up construction, but I don't think you have a clear explanation for who the users are. They must be:

1) Sure they are only going to use models that are supported by NN-512. This is a big jump for me and many people who want to do transformers, LSTMs, etc. and not just what NN-512 costs. 2) They have to be price sensitive. But if you want larger companies, they will rarely just have one use-case that NN-512 supports; they will probably have a ton of use-cases that NN-512 supports. 3) They want to target AVX-512. If they are a medium or larger company they probably want to have tooling around this, to support model development and later deployment.

Those are my thoughts in general.

I currently need to be able to target and deploy on PC and Mac CPUs from the past 10 years. (I know, this is specific.)

I think parent is saying that individual devs are more likely to try NN-512 if they don't have to learn a new API, since most of them want to have the flexibility to fool around before committing to a new tech or bringing it in-house.

I guess my suggestion is really try to specifically define who the user is for NN-512 and what NN problems they are trying to solve, and let that shape the roadmap.

My original thinking was that companies needing to use transformers, LSTMs, etc., would want to fund the development after doing the simple work to evaluate NN-512's ResNeXt or DenseNet performance on their hardware

Same with development of tooling (network conversion, parameter import). It just takes a lot of time/effort for me to write the code. It's easy to say that support for AVX, AVX2, etc., would be nice, but that's months of hard work (unfunded)

I guess we agree that a niche tool like NN-512, that has a reduced feature set (trimmed to allow one person to complete it and release it for free), has trouble competing with the huge, free, flexible, multi-target tools funded by large companies

In retrospect this is obvious!

TVM can do that for you: http://tvm.apache.org/

Just wondering, what makes NN-512 superior to Intel's oneAPI DNN library?

For example: try to find a deep learning framework that uses Intel's oneAPI DNN library and achieves 80% of the performance of this: https://nn-512.com/browse/DenseNet121

Tell me if you can find one

I had a similar idea last year, but clearly I'm not talented nor experienced enough to do it. Congratulations dude, this rocks.

Your project is seriously cool, I'm sorry it didn't work out for you.

Thanks dude

Where did you land?

If I remember it correctly, I initially wanted to build a LinkedIn alternative; but then it morphed into a IMDB alternative to give credit to where it's due


Besides answering who built what https://theymadethat.com/things/k4z/iphone

and who worked together and what makes up what https://theymadethat.com/projects/7da659d8-629f-5c10-9160-7a...

You can also use it to figure where something was used https://theymadethat.com/things/3r1/storyboard/show_used_for...

Different versions of a product also have their own profiles https://theymadethat.com/things/6wy/apple-macintosh/show_ver...

I'm still maintaining it, with some modest future goals of making the UI more mobile friendly

This is cool. Are all of the entries manually added, or are you also pulling in information from outside structured sources like Wikipedia?

Thanks. Data was manually added to dogfood the UX. It's improved a little from that.

Programmatically importing data is actually my next "big" goal. It'll be hard to maintain quality though when I start on this route.

I and a few classmates in business school built a service to print people's top 5 instagram pics (top = #comments + #likes) every month and mail it to them as a nice physical keepsake. I was really flabbergasted that we couldn't get even 10 customers, we spend nearly $1,000 on ads before calling it quits (if we had had 10 customers we may have kept iterating but the setback was big enough to decide to just stop, and it only took a few weeks to build). Part of what was happening at the same time was the move towards stories, videos, and posts with multiples images, so it would have become way more complex than initially planned anyway and none of us were engineers. The year before that I "invented" a mouthwash that was all powder based so you could travel with it despite travel liquid restrictions, the kickstarter campaign raised a decent amount but about a 1/3 shy of what it would have cost to place the minimum order with the packaging company (the power would be in tea bags), so I also nixed that but that was a fun side project

I’m currently subscribed to a service that allows me to select 8 pictures every month and have them printed and sent to me. They supply new albums at extra cost when the current one is full (have 3 so far, over about 1.5 years).

Apparently there is a market for this kind of thing. Mine is targeted at young parents though, maybe that makes a difference?

My family uses famileo.com (Europe only I think) which is basically that, selecting recent photos and add texts to make "articles" and they send your "family newspaper" to your older relatives each month.

My grand-parents love it, so yeah there's a market for this kind of things (at least as long as this generation lives).

We use famileo too, it's great. Our idea with that side hustle was total automation so people don't have to even choose the pictures, and perhaps that's where the product market fit was lacking. The hypothesis was that 1. most people think they'll choose pics but in reality they rarely do and rarely print any, and 2. in the "choose and print" market there was too much competition to begin with, from printing for pennies at walgreens or online via shutterfly, to recurring services like famileo

Here's the video I made, for what it's worth: https://youtu.be/93XUnsrG18g

curious, have you had any successes ? (respectfully)

I have, I built a subscription coloring book business when they became a huge fad. It was profitable (between $500 and $1,000 a month in net profit) and I sold it after 10 months when I needed to relocate outside of the country for work so the burden of managing it would no longer be worth it. I also joined somebody else's side hustle in the employee engagement space (tips for managers and a simple tool to keep track of staff info like birthdays, favorite food, etc.) that was acquired by a larger but still very small company in that space. Both "exits" were not in the low 5 digits so nothing notable but a nice feeling

What did you sell it for if you don’t mind my asking?

Chroma Club (the coloring book one) was just $3k, I had about 2 weeks to unload it because I didn’t want to have to handle closing and refunding subscriptions (some people prepaid for 6 months etc.) and it was faster/easier to sell than unwind, so a bit of a fire sale. The other one was about 10x that amount.

My German library app to track books borrowed from a library: http://www.videlibri.de/

It was doomed from the start, because basically no one goes to public libraries anymore. But you cannot go to a library without it. Like this week I saw a post of someone getting a 150€ late fee bill from the library while having no idea why, that would never happen with my app. That should have been well monetarizable, pay for my app, or you are going to pay ten times more to a late fee bill

After the Windows app failed, I doubled down on the niche, by making it a Linux app.

I thought as plan B, if I do not find users, I could sell it to the library itself. I thought that was a sure sale, because the library had bought software from me before, but they did not want an app

Then I made it open-source, so I do not have to deal with it anymore, but no one understands Pascal code. It only lead to much more work. Now I have been working on it almost every free hour for 14 year. And I really should not have written my own HTML parser for it

I haven't been in a library for a long time, but in 90s I never had this issue. We had a "card" where librarian would stamp the date. Books had a paper on the back where they would stamp the issue date and return date.

It was trivial to know when you had to return a book. Is this done differently in Germany?

I like the app but I believe people won't buy/subscribe to an app that helps you only if you forget to return a book in time. Most people will believe they will return it in time and won't invest in this type of "insurance".

Libraries might be more interesting customers but they have little reason to reduce late fee bills as it is extra income to them. You could/should look whether university/free libraries are interested.

Hope you enjoy working on it.

>It was trivial to know when you had to return a book. Is this done differently in Germany?

It should be. The due dates are all listed on the webpage

And some libraries even send reminder emails before the due date

But that is how they got me. I got used to their reminder mails, and then - "Our mail server had a failure and now you have to pay 60€"

>I like the app but I believe people won't buy/subscribe to an app that helps you only if you forget to return a book in time. Most people will believe they will return it in time and won't invest in this type of "insurance".

I wish someone had told me that before I wrote it

But there is also the secondary feature of keeping a record of read books. I read a lot of incomplete series before I had it, and now I want to read the rest of the series, but I do not remember what the series was.

>You could/should look whether university/free libraries are interested.

They are the worst. They have much less patrons than public libraries, and almost never answer emails. I think I might have gotten email PTSD from them.

>Hope you enjoy working on it.

I enjoyed it 13 years ago. But not anymore.

> because basically no one goes to public libraries anymore

Is this a German thing, an urban thing, or really a true statement universally? I ask because prior to covid lockdown, we made at least weekly trips to our public library and it always had a steady stream of patrons.

Compared to almost anything else not involving libraries

There are more people reading books than there are people reading books from a library. If I had just made a tool (in English) to keep a list books without involving (German) libraries, it would probably be much more successful and only take a hundredths of effort.

And worse, I did not target libraries in general, but libraries in my city. Anything limited to one city has a too low user base. I planed to expand it city by city, by selling it in one city to cover the travel cost to the next city, so I can lend books there, actually test it, and make sure everything remains bugfree.

The alternative option, expert users can use it with any library, by editing the config XML files to include their webpage URL and server parameters, seems to have been too confusing for people

> basically no one goes to public libraries anymore.

I do. This surprises me.

I always think a weekly trip to the library is good for the soul, just to be reminded of all the knowledge that's out there waiting to be discovered. Good reminder to watch less Netflix :-)

In my locale, the public library has shrunk in size to about 40% of its size 20 years ago. And it now includes a coffee bar and lunch café. The current selection is mostly popular books. For all topics I know something about, the non-fiction books seems to cater to children, lay-people, and beginners.

The library has pivoted to social work, helping low-literate people, offering a safe space for teenagers to study, and a place for the elderly and 30+ women to meet. Those are all good reasons to exists, particularly because many of the visitors cannot go anywhere else, but I miss the library of my youth. I loved spending hours moving through the stacks, discovering all kinds of oddities and new things. But that seems passé. Even in the university library the amount of books has shrunk a lot and the depot with a lot of older stuff is now off-limits.

The current trend in libraries is to abolish fines and waive any past fines. It has been shown to increase usage of the libraries.

Do people really not use libraries in Germany any more? In the US, they are more popular than ever.

> Do people really not use libraries in Germany any more?

I am really chuckling at the double meaning here.

> And I really should not have written my own HTML parser for it

> But you cannot go to a library without it.

Why is this? Did you arrange some kind of exclusivity deal with libraries in an area to manage borrowing so that people have to use it?

I think GP means that people should want to use the app because it can save them from having to pay late fees, not that people need to use the app. Like saying "Yelp, cannot go to a restaurant without it".

I am fascinated by hardware with poor software support so I built two highly unsuccessful native mac apps: MASS and LabelScope to work with weight scales and label printers. You can read their defunct websites at https://semireg.com/mass/ and https://semireg.com/labelscope/

Both worked in novel ways. MASS “floated” the current weight above all apps and when double-clicked would insert the current weight at the focused cursor point via keyboard accessibility APIs.

LabelScope printed to label printers but wasn’t a design app or a driver. Instead, like MASS it floated above other apps as a “scope” which would real-time capture and dither the image inside the virtual label. Double-click would send the dithered image directly to the printer using the printer’s native commands via USB (no driver).

Fast forward a few years and I started on ANOTHER label printer app based on Electron. I’m happy to report it’s a healthy business and growing! Learn more at https://label.live

Can you please please PLEASE support Brady printers? They're specialised label printers used in the laboratories. The software they use called LabelMark is no longer supported except lots of labs still have label stocks. So they're stuck used the pos software. To make matters worse its not supported on win10.

I'm happy to be your lab rat and customer for life because we are never going to stop using the printers. I've completed the survey on your site. Shoot me an email if you want any more details.

Btw every lab I've been in has used Brady printer

Replied! Thanks for the suggestion.

I had a similar idea for a web based version about 8 years ago. Even ran some Google Ads with a MVP landing page to gauge interest (there was).

Finally never got around to building it.

Is this a sustainable business for you now? Or just a side project?

Yes, it’s ramen profitable, bootstrapped and growing. It’s just me at the moment and I’m still juggling full-time mobile app and IOT consulting. Since Label LIVE is a React app wrapped in Electron I’ve given a lot of thought to bringing it into a proper SaaS via browser but just don’t have the time.

Nice. Congratulations!

A suspect a browser based SaaS version will knock the socks off of what you have now in terms of conversion.

Try and find some time to build it.

Wouldn’t it be hsrd to get the browser app to play nice with printers?

Yes, it’s a big challenge. There are three solutions I see:

1) Rely on PDF and browser print APIs, which means you give up direct USB access and even some paper size API which is kind of a deal breaker for thermal printers.

2) Rely on a browser extension of some sort. This might work, and I haven’t researched this much, because I find it clunky and error prone, bad UX and at mercy of browsers.

3) Use the existing app as an “agent” that receives “print jobs” from the cloud. This would be similar to ShipStation Connect and Google Cloud Print. This has potential but would need all sorts of engineering and UX to become seamless.

Could you say more about why you think the original mac apps "failed?" Small market? Lack of promotion? Cost too high? What?

Oh, definitely due to a small market WITHIN a small market. When users acquire speciality business hardware they don't reach for macOS. I'm trying to change that... one device at a time. MASS and LabelScope were mostly proof of concept. They failed in the sense that they were the cost of the education to do better. With Label LIVE (my newest app), I built an actual design app, targeting a much broader market (cross-platform to boot).

I was thinking about building the same kind of native Mac app as MASS and I never heard of it... my bad. Did you have prior USB programming experience? How did you learn to interface with the scales? Did you reverse engineer them? What’s the best way to learn to do this sort of USB programming on the Mac? Thanks in advance for your answers.

TLDR: Small wins over the course of months and months of trial and error.

I never considered myself a hardware integrator until I accidentally became one. Oops, e.g. fake it until you make it. Looking back at college, I received barely passing grades in my C and assembly coursework. Sure, I "understand" bitwise operations and bit shifting and LSB/MSB... but it still feels foreign to me, like every time I got something to work I celebrated that I made any progress at all. Me the imposter.

There were times in developing all these apps that I seriously thought, "well, I'm stuck, this is the end" and I just kept working a few hours at a time... If anything, troubleshooting and not getting fatally-frustrated is my super-power. These hard-won-solved-problems led me to build something "simple" for the user, which is its own very special reward (aka dev dopamine and serotonin).

I loved building these apps. I never thought I'd make any money. And my relative success with Label LIVE is just a rewarding icing on the cake of enjoying this kind of puzzle-solving. I mean, native Node modules, WASM... are their own special kinds of hell. The kind of hell you forget quickly and you have better written your future self decent documentation. Makes me nervous just thinking about finally adding Apple Silicon support. hands sweat

1990 Refilling toner cartridges for laser printers. I had someone else doing the refilling for me, so in terms of execution this was pretty much pure sales calls and I found that I wasn't good at that.

1992 A 1U rack mounted utility component for pro touring musicians that included a direct input (DI), tuner, metronome, front and rear rack lights, and surge protected outlets. NAMM attendees loved it. So did all the musician's that I worked with on the road at the time. But it turns out there's a lot to building physical products, especially ones that include custom electronics. We got a PO for 25 units from Guitar Center, but didn't have the chops to get it together and deliver. In the end it came down to capital and we had none left.

2000 A SaaS solution (then known as an ASP) for project management for commercial construction. Not sure what went wrong here. Probably lack of sales competency. A peer and competitor at the time, eBuilder, just sold for $500 million 2 years ago. That was a long ride. We shut down in 2008.

I think your 1U rack unit would still be something a lot of gigging musicians would want. You could probably add in a few other things to that space too.

An app for finding the best hotels for traveling with kids.

It was basically crawling the booking.com API and then applying an algorithm to figure out if a property was suitable for families with children or not. It would do the usual. Parse descriptions and try to find keywords there, parse and count the number of comments mentioning family-related topics, up-score hotels with certain facilities and then push the images through Google Cloud Vision APIs to re-order the images putting first the ones with children, pools, playgrounds, etc.

With that done, just an app to find hotels by searching anywhere.

It was more built for personal use but ended up being quite attractive. But marketing apps is hard so never really pushed it. Then, also... covid happened.

Rumour is due to covid a babyboom is coming, so soon there will be a lot of potential for your app.

There might be a baby boom after Covid-19, but right now, we are expecting a baby bust. Economic uncertainty is leading people to delay having children, at least in certain countries.

Had a similar idea to mine the booking.com data for cheap hotels. Kinda like Scott's cheap flight but for hotels.

I started the project but life got in its way.

This pandemia really crush all tourism industry...

For at least two years

You can share it here! You have our attention :)

this is a good idea. Keep it up.

People are starting to travel, atleast in India. There will be a surge in tourism as soon as the vaccine reaches its potential.

Most people in my circle are dying to travel. So stick around man

https://wifimask.com - WifiMask VPN. The little money that comes in is spent on server costs basically. I spent 3 years building it, spending too much time on trying to build "perfect" Apple apps, instead of quickly going online with an MVP and validate first. Besides that there are a lot of VPN's out there and competing with the "big boys" with big marketing budgets is hard, especially when they can provide things like Netflix US streaming. I can keep it online forever this way, it's "bootstrapped", so no screaming investors who want to see money, but I stopped working on an Android and Windows app, I see it as a lost case and these days I'm focused on my webhosting: joostwebhost.nl and secret future projects. ;-) (And I'm available as a freelancer) I learned a lot from it though, more then I ever learned at an employer. I'm pretty sure I can apply the knowledge to a future successful project.

The whole internet's full of VPN service but there's still a big market for it. There's a service I used a couple of years ago that allowed me to basically connect to the internet through this particular VPN without having any data plan. I don't know how they did it but it still seems weird that the service worked this way on the contrary to what I understand VPNs actually do or how they work. After a little bit of googling, finally got the name it was called Your Freedom

Interesting, this was probably due to "DNS tunnel mode": https://www.your-freedom.net/index.php?id=dns-tunneling I will look into this more, thanks! :-)

I guess based on https://github.com/yarrick/iodine ? You could also run your own server.

Good on you for sticking with it. I too tried launching a similar service in this market, you’re right, the “big boys” marketing is everywhere and drowns out smaller service providers. Keep going, good luck.

Thank you!

I bought lacktivity.com with the hopes of building something that would make fun of people for being inactive so much. Then I looked into what it takes to build a mobile app and gave up. I'm not a dev. But it was going to feature such gems as

"Congrats, you've been inactive for 4hrs 32m: A new record!"

"Lacktivity has crashed due to sheer boredom."

"Wow, 100 steps in an hour! Next time, try taking the longer route to the fridge."

"Putting the fitness tracker on your dog and then playing fetch in the house is no way to go through life."

It was a fun idea, but I totally lack execution.

Domain for sale if anyone wants to pick up the idea.

This would have made bank on iPhone back in 2008 when novelty apps (eg BeerMe) were big.

I made a baby names app in 2010-2011, mostly driven by my and my wife's inability to settle on a name for our first child. The system learned n-grams from names you liked, and other info like name origins etc. I tidied it all up into what I hoped was a tasteful Mac app, used some free Google Ad credit, and it was very satisfying to see that people used it, but it was never popular enough to pay its own way, despite briefly being the number one lifestyle app on the UK Mac App store.

I would have done better if I'd tweaked the algorithm more, done an iPhone version, or just taken it seriously, but it mostly came out of a personal need and in the absence of clear signs it was valuable it was easy to let go (especially when later sandboxing rules for Mac apps broke some fundamental stuff around persistent documents).

You can see some screenshots at:


I've spent a whole bunch of time on CoolBeans (now OpenBeans http://www.openbeans.org), a "distribution" of NetBeans.

I assumed it would be possible to do a lifestyle business out of catering to a subset of the over 1M NetBeans users. Turns out it is not so easy.

I spend some money for a Windows machine to digitally sign builds, the macOS dev certificate and hosting. But the bulk of the cost was the time it all took.

PS: The .xyz domain in combination with the Windows antivirus solutions (looking at you ESET) was a major annoyance. Switched to .ORG just in time for the whole debacle there.

Why didn't you get a .com address? Reminds me of this oldie:


Seems obvious in retrospect. I didn't believe the name is so important, CoolBeans was nice enough and .xyz was being promoted as the new .com :)

Back in the early days of Flash, I realized it could be more than just for cute animations, and coded a library of UI widgets to enable Windows-like UX inside of a Flash app. I then made a sample CMS using it that let you drag/drop panels and widgets into a page and it spit out HTML, so non-designers could design web pages.

Sadly, the next version of Flash showed that they were thinking the same thing - all those UI widgets I made were built-in on their next release, which came out the day after I finished my first version.

Lesson learned - know the roadmap of the products you are consuming.

I made a flexible, light-weight, card-based drag-and-drop kanban-ish planning tool called StoryWall and posted the MVP on Hacker News literally the same week Trello was released.

Managed to take all the wind from my sails.

StoryWall name sounds like an outliner app, maybe you should have pivoted into that.

My wife has a side project selling antique maps. I built her a shopify site about 1 year ago that costs €30 p/m but it's hardly generating any revenue. I spent maybe 10 hours on it and she has spent about € 2000 on inventory. https://utrechtaandemuur.nl

This looks good! There are some minor UX tweaks that you should do for a better starting point to the site. I'm assuming:

1. The initial site hasn't changed since launch.

2. If it has, then its been due to intuition vs data.

From my quick glance, this has potential and she should be able to sell her entire inventory €2000. It would be good to know what the digital marketing strategy has been to build awareness and drive sales from her target audience as well as a better understanding of her customer and industry.

In any case, she has something and I hope she doesn't give up.

Nuce business. I didn't see an English version of the site.

Is it accurate to assume the target audience is only Dutch-speaking people? This is cool but I haven't wanted antique maps of the Netherlands specifically.

Not just Dutch speaking but looks like it's specifically the city of Utrecht. Which is pretty specific, though the hostname tells you as much if you speak Dutch (it translates to 'Utrecht on the (your) wall').

Wondering: why did you choose to focus on one city?

Get the prints into retail spaces - book shops and boutiques.

Why not switch to WooCommerce and I host it for you for €49/year. ;-)

have you considered localizing and selling it to english speaking countries too? cool looking maps but not all that accessible.

This is very niche. I question who would really want to buy antique maps, except as for wall art.

Can you venture into other things to sell on Shopify instead?

If you were the world leader in antique maps, you’d make a pile of money from that niche. Being able to search for authentic maps for anywhere in world, sorted by time period — that would be a big hit in certain circles. You could ask who would buy first edition printed books — but that’s also a pretty large niche with lots of passionate customers. For example, I’d be interested in old aviation charts or wartime maps from World War II or the US Civil War/etc.

It's not as niche as you might think. I think the problem is that there are already lots of other websites selling antique maps. For example:

- https://www.raremaps.com/

- https://artsourceinternational.com/

- https://www.swiftmaps.com/product-category/antique-maps/

Store Locator as a SaaS. There are some successful products in this space, I just didn't execute on sales.

User created API consuming dashboards, I drove lots of traffic to it but it was a really niche product and I just don't think there was a path to revenue with such small numbers.

Static pagebuilder before Square and the like had that kind of thing dialed. Still proud of the technical achievement, had some users, but it never took off. I couldn't afford to market it.

A more traditional digital agency I started with some colleagues. It went well, and then it fell apart when one member grossly mismanaged the money and then took some big clients on his way out. I regret not sticking with it, but it was really demoralising that such bad behavior was rewarded and we were left with nothing. Many lessons learned though, I would never allow it to happen again.

Many other small projects that just waned or never got built to completion. I do all of this on my own and with my own money outside of any tech bubble, so it is pretty easy to get derailed. I have been at it for more than 10 years.

What did store locator do? We’re hosting costs too much?

I made a tool called Diskache, which would combine an SSD with an HDD into a hybrid drive around 2015. You did not have to use the entire disk(s) for that. It worked on any two drives on Windows. But my marketing skills got me 0 customers beyond myself.

The target audience was supposed to be gamers (back then typical SSD was 150GB and GTA V was 55GB alone). I even got greenlit on Steam: https://steamcommunity.com/sharedfiles/filedetails/?id=61003...

Hybrid SSDs were available in 2007, Apple's Fusion drive in 2012, along with ReadyCache/ReadyBoost/ExpressCache/ReadyDrive (remember when some laptops came with a 32GB SSD in addition to their HDD?), the number of people willing to pay what this product is worth, but not just buy a bigger SSD instead, is probably very slim.

Also tail end of a niche but long flooded market. I wouldn't beat up your marketing skills too badly for it I the Steam page probably would have gotten a decent number of conversions if the demand had been there.

Hm. At the time of the release there were no alternatives not bound to specific hardware (e.g. Intel RST on specific chipsets). I am not even sure there are now. AFAIK, Intel RST is dead, and AMD has no alternative either.

I built https://pagewatch.dev (a service to test your site for responsive/layout issues) over the last year. I have a few customers but it is growing much slower than I expected. If I were to start building it now I would spend much more time building some kind of audience, just driving even a few visitors a day is hard starting from scratch.

I spent about 6 months building Itinee (https://itinee.com) which is a trip planning app focused on being budget conscious. I built it because my wife and I love to travel but my assortment of spreadsheets was a little intimidating to her. I wanted a platform that let us both participate in the planning process. Unfortunately, due to COVID, people aren't traveling so I haven't really put any money into marketing it or done any more development on it. I might revisit it once things get back to normal. I don't plan to do much more development until I get user feedback, though

Looks quite interesting. I'm curious if market product fit is there, you are trying to sell an app to people who are looking to travel cheaply.

Does your app offer some kind of incentive for use. Perhaps utilize data from other users to suggest how much certain city/trip should cost. If you need data for that, it might be a good idea to offer the app for free/freemium. Perhaps for free you can do planning but with premium you get further insight for average cost of a specific trip or even suggestions for trips of various types/cost.

My grand vision is actually a "trip itinerary search engine" so people can search for "Barcelona for 2 people for $1000 or less" and get all the matching itineraries.

But yes, as you so rightly pointed out, such analytics is only possible with a user base, which I've been struggling to acquire (admittedly, I haven't tried very hard). A free model may be a good way to go to build that user base initially, and may be something I explore, but I didn't want to give something away if people WOULD pay for it.

One thing I've considered, if I do go down this route, is that free users' trips are public (or mostly, I would try to censor dates/residential locations) while premium users can make private trips or something like that.

Overall, I've gotten a lot of feedback on this thread to reconsider the pricing/monetization strategy, so it's definitely something I need to look at.

How'd you come to the pricing model? To me only having 24 hours before losing data would probably preclude me from putting effort into trying it out. Have you thought at all about a freemium model?

I gotpartway through building a travel planning app a few years ago, and I had thought the best way could be to allow 3 trips to be planned, then your trial ends unless you subscribe. Or you could severely limit the functionality in a perpetually free version while the paid version has all the bells and whistles.

All the best for when travel is common again as I really like this idea and think there’s untapped potential.

Yeah, the issue I see with something like that is that realistically, someone could easily take several years to take 3 trips.. seems a bit tricky to convert free customers to paying customers.

Thanks for the feedback and well wishes!

Pretty much by waving my finger in the air :) I originally started with something subscription based, but it felt a bit inflexible. For example, if I offered 6mo subscriptions, people probably wouldn't purchase a subscription if they were planning a trip to take in the next month, because you're "wasting" 5mo.

Now, it could be that my prices are too high, but that was sort of my rationale--I thought it would be better to let people pay for what they need.

I think a freemium model could work, but at this point I don't think I have any real features that would meaningfully distinguish the free version from the paid version. I did consider limiting like the number of days your trip can be for "free" users--I might revisit this in the future.

Really appreciate the feedback!

I would be the target audience, but from the front page I don't see what value it brings over a spreadsheet. I won't register just to found out that. Does it automatically calculate travel costs?

Thanks for the feedback! In my view, these are some of the biggest advantages of Itinee over my previous spreadsheet planning:

- Visualize everything on a map. This is big for me just to help organize what days it makes the most sense to do things - Estimate travel costs (e.g. Ride sharing) based on distances between stops - Easily adjust number of attendees. Some costs are split regardless of how many people attend, like hotel rooms, while others are a fixed per person price, like event tickets - Ease of use for someone NOT familiar with spreadsheets. My wife couldn't replicate any of this in a spreadsheet, but she can use Itinee.

Some of the above may in fact, and probably is, possible with some fancy spreadsheet shenanigans, but the main point was to make the whole process more accessible.

I'd be really curious if the advantages I mentioned weren't apparent on the website or if they simply weren't, in your view, "enough" of an improvement over a spreadsheet to justify paying for it.

Estimate travel costs weren't apparent from the front page, there is no mention of it and the example plan (picture) only use walking.

I just found the Chicago itinerary in the blog post. It should be accessible from the front page.

I don't see enough value added over spreadsheets / custom Google maps to worth the money, but I'm pretty happy with our current system.

This is great feedback! I definitely think it's a harder sell for someone who already has a system that works for them.

Thanks again for taking the time to respond. Happy new year!

Interesting concept, although the pricing structure seems at odds with budgeting. I would say 6-month or 12-month subscription blocks make more sense to allow for planning and budgeting.

Thanks for the feedback! Pricing is not my forte. My thought process was basically this:

- In general, people are probably only planning a single trip at a time (an exception might be if travel agents were to be interested)

- Based on that assumption, it came down to what resolution of access people would want. With the current model, 6 months of editing access comes down to $35. I could, equivalently, just charge $35 for 6mo of access, but in my head, I thought people would be more likely to actually give it a shot if the barrier to entry were cheaper (in this case $10).

So it could be that my actual price is just too high (i.e. $35/6mo of editing access is too much). I haven't gotten much feedback one way or another on this (though one could argue that the lack of purchases could be seen as pretty clear feedback). I've just read that founders tend to underprice their SaaS services, so I was wary of starting too low.

I'm coming at it from someone who does regular budgeting and subscribes to software for that. Big communities built around that so there is potential also.

The pricing comes off as aimed at those planning big one-off trips not regular annual vacation. Those one-off trips are nice, but would be more sporadic and wouldn't you prefer a regular annual subscribing user over a one-off?

It generally costs more to gain a new customer.

At $35 for 6 months, that would be $70 annually if I wanted that. As what I'd consider an "add-on" product, it's far too much. Maybe something like $4.5/m or $45/yr.

Yes! Budgeter here as well :) My YNAB subscription is well worth it, BUT I use YNAB at least weekly. That was a big struggle I had with coming up with a pricing model that made sense here--realistically, it seems like most/many people will only be actively using Itinee for relatively brief periods of time throughout the year, which, at least if I were the customer, would make me wary of long subscriptions.

As a developer/founder I absolutely would prefer regular subscriptions to sporadic one-offs, I just doubt(ed) if people would actually signup for that.

I certainly agree that transitioning to a subscription model would require lower prices to be feasible and may be sufficient to get people on board with "wasting" their subscription for 75% of the year.

As a tangent, I find that it's really hard to get this kind of feedback from real/potential customers (like yourself!) but such feedback is incredibly valuable (thank you!). I wonder if there's a market for some service where I could pay $100 for 3 people to go to my landing page, try my product (free of cost to them, obviously) and then be open to some conversation about their experience. It seems like there could be a market for that given the "Indie Hacker" boom. Maybe such a thing already exists.

How about something like a quasi-log scale, e.g.,: $19 for 6 weeks, $39 for 6 months, $79 for 12 months and $99 for three years?

That could work! I'm curious what makes subscription models more attractive from your point of view. Is it just simpler to understand? It seems strictly less flexible than a pay-for-what you need model. I suppose in your example the advantage is that you can get a discount by paying for more up front which is a clear advantage for people who will take multiple trips in a year or are planning a trip far in advance.

I received a similar suggestion years ago from someone else, and it worked very well. It's effectively bigger discounts for longer & larger service plans up front. I found it effectively shifted the conversation from "How can I get a discount /beat you up on your price?" to "Ok, what is the right plan for our project"?

This is not quite like consulting/custom dev gigs, but has similar attributes. For purchases/subscriptions like yours, I've noticed that they often seem to create a tension between buying the two most likely to meet your needs - the longer higher-cost one is priced just low enough to be tempting. This probably indicates a lot of A/B testing of pricing, and decisions about whether you want to make it easier for your customers to decide, pointing customers to one plan or the other, what really drives higher purchasing levels and overall income.

Good luck!

They're suggesting a pricing model from the perspective of running this business instead of a pricing model from the perspective of using this service. Subscriptions are worth way more to a (tech) business than one-off purchases.

> To save your changes, purchase a trip slot for as low as $10

Yet the price directly to the right of that says "Trip Slot $5".

Thanks for the catch! I updated the language to "...purchase a trip slot and editing access for as low as $10"

Your comment makes me wonder if it would be clearer to present things as "Trip slot + first month of editing access = $10" and "Each additional month of editing access = $5" or similar.

Ah, I see, that makes more sense. I had not realised that a "trip slot" was a read only thing although on reflection that's pretty obvious. I thought the monthly price was unlimited slots...but that makes no sense given it's the same price.

During the previous crypto craze a few years ago I built a .net service that continuously scanned 100s of sites for crypto news for specific keywords (configuration of sites and keywords stored in xml files) and texted me alerts when keywords hit, in an attempt to apply an event-based investment strategy to the crypto space. Given the immaturity of the market, small events often led to irrational spikes, and I made some money on quickly acting before the news became too widespread.

I started getting too many alerts so I built a machine learning based system to weed out uninteresting/unrelated stories (basically a spam filter, was 98%+ effective). By this time the service had a database of 100,000+ headlines for crypto stories, so I had good data for the filter. I built a UI to quickly allow me to train the ML algorithm by manually rating stories as good/bad and give me the ability to train/retrain/enhance the models on an ongoing basis as my crypto interests changed.

When the market crashed I lost interest. Was already full time on something else as well. I had visions of deploying the system publicly as a SaaS product, allowing people to sign up, and replacing the config/xml files with a nice web-based UI, and then charging something for it.

Why didn't you use the Google service (sorry can't remember the name) that warns you when your keywords appear somewhere on the web?

I didn’t trust that it would be fast enough and don’t think it could send text alerts. Minutes counted in these cases. Also architected it to plug in any source, not just websites (tweets, discord, dark web, etc.).

You do realize that the bull run is back again? Might be a good idea to look into this again.

Yeah thinking about it! It was time consuming. Also I’m thinking the market this time around will be more mature than last time but don’t know, watching it.

I would like to use this. Any way to get access?

There isn’t at the moment. Curious what part interests you?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact