Ask HN: What (side-)project are you working on?

101 points by chdaniel ↗ HN
It's been ~one year since the first lockdowns started. What are you working on now?

I'll start: I'm building PriceUnlock, a tool that will help you find the best pricing for your SaaS product. I'll post more in the comments so as to not hijack the post's description!

243 comments

[ 2.5 ms ] story [ 324 ms ] thread
I'm announcing PriceUnlock here: https://bychgroup.com/price-unlock/

But the gist of it is: * We're aiming to make it easier and less risky to try charging more/less * Aiming to help you stop leaving money on the table (or take too much money from the table?) by enabling you to easily charge different prices in higher/lower-income countries — like Apple, Netflix, etc. do * Aiming to make A/B testing of prices headache-free

I'm still building PriceUnlock and revealing more bit by bit! Curious to hear your story about your SaaS pricing and how you got there — and whether you ever got to 100% confidence that the prices you set were "right"?

That aside, would love to hear what others are working on now to see if my feedback can be helpful!

Interesting, I think there's a lot we could improve in the way we do things. You're talking about an automatic pricing tool that learns the market through time and region and automatically seeks optimal price points?... Some kind of drop in library tool a few lines of code thing would be awesome! Also if you're making me more money, I'd be willing to share a percentage with you considering I may not have had that money otherwise.
Wow thanks that helps a lot!

Not really automatic price-setting, at least not by default! I would personally like to have control over the knobs in my product.

It would give you a dashboard where

A) You would be able to set the prices manually (with suggestions, of course, and with taking into account things like extra Stripe fees, rounded estimations so you don't wound up with £28.837 in the UK).

You'd set them manually based on country, so you won't leave money on the table by not charging less in India, for instance.

I'm always mentioning India because Apple is the widest example that comes to mind: Apple one is £15 and £25 in the UK, whilst $1.5 and $3 in India

B) Run tests that you set manually on prices. $19 vs $29 — see, with metrics, which one brings your company more revenue. Or perhaps same revenue, but more users, so more word of mouth — who knows what the test will show?

You'd be able to either schedule (e.g. $19 for week 1, $29 for week 2) or run tests at the same time (show 50% of ppl $19 , show 50% of ppl $29 // OR show ppl in the Uk $19, show ppl in the US $29)

RE: your last comment — the best part would be that it'd be a flat fee per month, no percentages! So every extra cash that you receive from ppl on top of the flat fee is yours to keep for... well for as long as your "LT" in LTV!

How does that sound, you think?

So, essentially, you're building an A/B testing tool specifically for pricing?

Thea complexity you've alluded to in your various descriptions start to get really hairy once you scratch the surface, for example:

You mentioned trading off lower prices for more users. Whether the tradeoff is worth it depends not just on the difference in income, but on the marginal cost of serving another user (which itself may be variable).

You mentioned more users leading to more WOM. That is a huge assumption, you have to actually measure your viral coefficient (which might actually change with different prices) to know whether that hypothetical WOM actually leads to more users.

You may need to optimize for cash flow instead of just income. As one example, you could try a higher monthly price combined with a larger discount (two months free instead of one) for paying up front for a year. That may lower the total number of signups but a greater % pay for a year, improving your cashflow (and with fewer signups, you may have better profit margins on a similar income).

You can add an additional expensive plan to act as an 'anchor' that no-one ever picks but makes the regular plan seem like a better deal.

There are so many ways to adjust pricing, and every part of your conversion funnel (and business model) may be affected, sometimes with counterintuitive results. Figuring out what some change to your pricing plans is really doing for (or to) your business can be really hard. Instrumenting your funnel and business model so you can actually see what a pricing change would really do to the businesses finances over time would be extremely useful.

MultiPreview: https://multipreview.com

Create sharable links with different images, so you can share the same article multiple times, e.g. on twitter keeping your feed nice and clean.

For those interested in the details, it renders an html page with custom metatags and immediately redirects to the target. Redirect happens in js, so crawlers actually display our metatags vs the target ones. It's just a hack I've tested a while ago, a couple marketer friends liked it and so I decided to make it into a micro SaaS.

If you want to try it out, I recommend to have a blog post handy. (also, credit card are disabled, but I like to keep the landing page "final".)

Ohh, kinda like Seth Godin's blog! If you share any of his blog posts, there's a rotation of pics that come up in Twitter's preview. Did I get it right?

P.S: Feedback for feedback, if you're down? https://bychgroup.com/price-unlock/

Thanks - do you mean increasing the price or blocking signup with credit card? The first I can do. :) The second it's not worth my time to handle the complaints, especially if I post it on HN to ask for feedback.

For Seth Godin's blog, it seems he's periodically rotating the image tag at his blog level. All the post get the same image for a certain amount of time.

With MultiPreview you set the images you want. But the time rotation could be a nice feature, I'll add it to the list! Thanks again!

I have been working on SceneRadar.

This tool displays you timestamped alerts (nudity, sex, violence or gore). So if you plan to watch a film with your child, your parents or other conservative relatives - you know when the danger scenes are going to hit you.

Since I watched lots of films during the lockdown, I thought this would be a good use of time.

https://sceneradar.com/

This is a great idea! Are you doing it manually or scraping subs for keywords etc?
Thanks. I have been manually adding the alerts. Users also have the option of submitting their own alerts.
you probably know this but you're building a nice labelled training set for machine learning to help you automate the process later
I filled my own need with https://howsyourblank.com.

I have a minor, but chronic medical condition I am trying to get in check, and I just wanted something incredibly simple to identify good and bad days. I was inspired by "year in pixels" calendars.

And I took the opportunity to try out Userbase[0] and build something with a secure backend and no tracking, considering the potentially sensitive nature of it.

No plans for monetization at the moment. I could see adding more features such as tracking multiple data points, stats, correlations, notes, etc., and creating a premium version. I would need more users and feedback.

[0]https://userbase.com

great job! you took a problem, came up with a solution, and executed it very well.
I think this needs a demo page/account showing how it looks. Looks interesting, but I don't want to create account just to see how it works.
Agreed! I've added a couple screenshoots of the main views within the app. I know more can be done around this, but hopefully this is a good start. Thanks for the feedback!
this would be great for headache and sleep tracking, both of which I need.
Dotfilehub: https://dotfilehub.com

I've always found various solutions that use git for sharing configuration files cumbersome. I set out to make my own simple version control system, and a lightweight web application where I can browse and edit them remotely. The main idea is that paths are aliased to simple names, so I can say `dotfile pull i3` and it will install https://dotfilehub.com/knoebber/i3 to ~/.config/i3/config

Overall the project is stable and I use it daily for all sorts of miscellaneous files.

Would love to comment with feedback but I'm not techincal enough to use it! Sounds like a time-saver though. Do you have any plan with Dotfilehub? Sell it as a product, keep it free, accept donations if it becomes widely used...?
My plan is to find some users for it, but I'm not sure the best way to get it out there.

I don't think I'd ever want to accept donations or sell it. It should always be pretty cheap to host unless it blows up, which I kind of doubt. But who knows!

Great idea! I feel the personal configuration space is ripe for innovation.
Pretty cool. The kind of thing I’d never look for, but now want to explore.

I’ve seen some awesome looking terminals / setups while watching programming demos / tutorials.

I would totally try their setup if they had a link to it via your site.

Likewise, any sort of preview / screen shot on your site would be awesome.

Well executed thus far.

HTTPS://PDAP.IO a nonprofit with the goal of scraping and making accessible county level police data. With this data aggregated and made available, policing the police will be more possible.
AFI SWiM: https://afiswim.com

Extract crucial information from Air Force Instruction manuals (AFIs) and produce an Excel workbook with the results for better reference.

I'm building https://pygma.app, a tool to turn Figma files into real websites. Perhaps not quite a side project as I've been spending half of the week on it. Should be launching this week.
Sounds challenging to do well, but very cool concept. Best of luck.
Thank you. It certainly is challenging but it's mostly working.
I'm currently on Sketch, but been looking at Figma as I like the browser-dependency, as opposed to OS-dependency.

Question (and bear in mind I haven't used figma more than 5 min): what do I have to do besides the mockup in order to be able to use Pygma?

Looks v cool!

P.S: Feedback for feedback, if you're down? https://bychgroup.com/price-unlock/

You have to use the auto layout feature within Figma to have a consistent output for now (I intend on relaxing this requirement, but auto layout makes things much easier). But you don't need to do anything external to Figma. There's a plugin (currently in review) and you will be able to manage everything from there.

Will come back to this post to give you feedback later.

Hey, sorry I just realised I forgot to come back to this, here's some feedback:

I think the landing page is really well done - that kind of format can be bad if the content isn't well written but each sentence made me want to read the next so I think you've pulled it off. I do think some visuals to reinforce your message might help (e.g. a graph of willingness to pay vs location, etc.).

My understanding of this is that it's going to sit amongst your pricing code and run experiments to change the price dynamically and return the best result after some time - is that correct? If so, I do wonder why no one has pulled this off particularly well as a SaaS before? (or have they?). I believe that Profitwell for example makes their revenue from consulting.

The other thing is that if my understanding above is correct, what stops the user from just cancelling once you've optimised their pricing and you lose capturing a lot of the value you've created for them (again, you revenue model may be completely different, I don't know).

I would also question whether this kind of thing isn't ultimately solved by the old "talk to you users" thing and understanding what they would pay from that strategy rather than software (at least with purchasing power parity though software + independent research does seem more likely to work since it's hard to talk to people from all over the world).

Ultimately this is the kind of thing I would give a try once I have some moderate amount of revenue and I feel that I've at least figured out who my customers are and can then optimise from there.

Pretty cool anyway, hope it succeeds and I'm in a position to try it out in a few months.

(comment deleted)
Custom renderer on top of a classic game.
Coupling: language learning app for bilingual couples looking to pick up each other's native languages

https://learncoupling.com

Cool. I've fiddled with A-Frame and I've seen supermedium before. cool stuff. Why do you focus on couples though? I love the concept of partnering together but I don't think you should limit your audience.
Awesome!

I'm just starting with a clear niche and solving my own problems. Where many of my friends and myself are in bilingual relationships that have been trying to learn our partner's languages.

In the US, 15% of new marriages are multi-ethnic, so that's pretty large already. I think people are surprised how common it is. And it easily pique's people's interests who are in the situation versus another generic language learning platform.

If the project goes well, it can expand and branch out (e.g., adding friends/family, connecting to other peers, finding strangers to learn with). Right now, it works very well as a language learning tool, even without a partner.

I like the idea of learning a language with a spouse or partner, how about the usecase where neither person is a native speaker of the language being learned?
Yeah! I actually do have some users learning Japanese together with a platonic friend with the app. It seems to work well. The use case for learning languages with friends and other family is wide open, just taking the baby steps!
Like most people do I seem to start a million and finish none. I have been having trouble tracking my options trading cost basis lately and am thinking about writing something for that. I want to integration Plaid, but the version that supports Brokerage accounts starts at $500/mo :|.
I completely identify with this. it's about bad time management or whatever, no tool can fix this - it's about education. well organized people use excel just fine to track whatever, and also some artists don't track anything and do amazing. so it also depends on your purpose.

this is something I think about a lot. I think this affects some people more than other, not sure if this is ADHD, ADD or whatever but it is a huge problem even more so the information age.

let me know if you have ideas how to tackle this.

Daily tax parcel data (shapefiles) from 21 of 39 county's in Washington State.

https://waparcels.tax/

Shapefiles are sort of a rare format. Hoping to ingest all the data into a SQLite/SpatiaLite database to make it a more general purpose data source.

https://apps.binnyva.com/que

A web app that will give you a list of questions that can be used to break ice or know others well.

You can be with just one or more friends - once person goes thru the questions till they select one. Then everyone answers that question.

Software Delivery Simulator https://softwaresim.com

A Rust- and WebAssembly-based simulation product that helps design, communicate, and analyze systems of software tooling. It’s a fun and interesting side project, which should be finished pretty soon.

https://planflow.dev - The easiest way to PLAN your website or mobile app.

It tries to make planning a website or mobile app as close as possible to the pen and paper experience.

It does this in a simple but (from my, and the experience of those who have tested it out) very effective and engaging way.

I share a bit more about how this is done here: https://simpleprogrammer.com/information-architecture-develo....

Just subscribed, Tapha! Curious to see how it'll work, as I do UI/UX design as well

P.S: Feedback for feedback, if you're down? https://bychgroup.com/price-unlock/

Awesome! Thanks chdaniel. Happy to give you feedback. Check back in a few.
Alright, so what you have here is pretty innovative. The problem for me was that it just took way too long for me to figure out what it is you did.

You might be spending a bit too much time on the problem here. It may be worth spending a bit more on the solution and making that clearer.

Right now it seems that its hidden within the modal underneath the copy. It should be much higher and not within a modal.

I would also try to say more with less with the problem. Again, it felt like it took a little too long to get to the point (and the point you're making is a pretty good one I might add!).

I find that with people who are more on the 'founder' end of the customer spectrum, particularly those who are tech focused, too much 'copy' and 'slow salesmanship' can be a negative. We tend to err on the side of 'just tell me and I'll figure out if its valuable' if that makes any sense.

I hope this helps?

Good luck man!

Tapha

FYI - the landing page tells me nothing about what it is you're selling, and "Why PlanFlow" leads to "Access Denied". I don't want to give away my email if I've no idea what it is I'm signing up for.

Plus some bugs you might want to be aware of - the "Coming Soon" floating labels are pretty off-centre on the links, and the hamburger menu when the browser is in portrait is nearly unusable and splits the "Sign Up - Free" button in half. This is on Safari on a Mac.

Thanks man. Yes, we've had some issues on Safari for a while now (seems like its the new Internet Explorer :)). I'll take a look at this.

The site in general is in 'under construction' mode right now, so rest assured these issues will all be resolved by launch. We're just dedicating all of our energy towards the first version right now, so have not polished everything.

https://liet.io/ - A text & document classification API.

I'm focusing on legal text and documents over the more common "sentiment" classifications. Early days -- lots to do.

I'm working on a system hand off nodejs workloads over a serial port to another machine running some host software.

My intent is to easily write Internet-connected software for old machines where a host machine is doing all of the heavy lifting. I have been messing with 68k Macintosh systems first. The code is very much a work in progress that I am actively chipping away at, and not in a usable state just yet. I write a lot of nodejs professionally but haven't used C since college so its been a fun project.

nodejs software for the "modern" machine: https://github.com/CamHenlin/coprocessor.js

C software (targeted at a 68k mac) for the "slow" machine: https://github.com/CamHenlin/retro68-coprocessorjs-test

A StackOverflow / Discourse alternative: https://github.com/denysvitali/dev-portal-frontend https://github.com/denysvitali/dev-portal

Porting Linux to the Surface Pro X (and failing doing that due to my lack of kernel experience): http://github.com/denysvitali/surface-pro-x-linux/

Creating a Microsoft Teams library + client (WIP): https://github.com/fossteams/teams-api

Unfortunately my free time is quite limited :(

The screenshots here look interesting: https://github.com/denysvitali/dev-portal-designs

Any live demo?

Thanks!

Still WIP, no live demo at the moment, but the screenshots that you see on the dev-portal repo are already what you get :)

The design reference that you link is how I want it to be, I spent quite some time focusing on the look and feel with Figma, and I'm really happy about the results, I just need to do the 90% of the work now :D

We created https://www.getwelder.com/ as side project. Turn into main project and then back to side project haha.

Great for recording content (interviews) remotely easily & in high-quality.

This pic: https://assets.website-files.com/5e93b09565892b4fad16784a/5f...

Drives the point home! Would love to try it, as I've recorded a podcast with 13 episodes before: https://chdaniel.com/podcast

What made you turn this from side -> main -> side project? What are you working on now, specifically, on Welder?

Sorry for the late reply. Was swamped.

Happy to hear that Welder seems meaningful for you. You should definitely give it a try.

Btw thanks for letting me know what convinced you, we are actually in the process of updating the LP and this is valuable feedback.

So when covid thing happened we started on a few side projects as we though it's a great opportunity to build something that will help people in current situation.

We created MVP of app for therapists to do remote sessions, system for creative people to share their thoughts, Slack bot for devs to streamline their workflow and Welder.

Welder took of the most in our validation phase.

We decide to jump full time into it straight away as the demand was big. Few months in we already had paying customers, but the growth was not as fast as we needed.

We started thinking about VC funding, but finally it didn't feel right, we loved the product, but not as much. We didn't feel like it's the next unicorn. We would be lying to the VCs and we didn't want to.

So after 8 months we decided to put it back on side track as we had two options

1) VC capital and boost the generalistic tool as hell > Nope 2) Finding a niche, stick with it, understand it and deliver solution that is different enough from other generalistic competitors who raised (we didn't love it enough and we didn't feel big enough differentiator is there).

We actualy share a lot of this on our YT channel if you are interested https://www.youtube.com/c/CrazyMonkeys/

Hope this is helpful for you!

Visual cortex seen-image reconstruction model
Reminds me of that one House MD episode... crazy
(comment deleted)