Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

My app failed: lessons learned

rithm.app

Post reply on HN
Print view View original

171–177 of 177 posts

Re: My app failed: lessons learned

#171

Post by asimjalis »

kylebolt wrote: post #100
bastawhiz wrote: post #73

> I initially saw oh ok, there are a ton of people making these apps, it must be profitable. If you look at a dense market and decide to join in because you see potential profit because the market is dense , you're already destined for failure. Everyone else already has a head start, which means you need some sort of differentiating feature to be competitive. Why would I pay for blog software that's half baked or doe…

I thought habits might be an interesting market to work in + using a chatbot as a novel approach but alas my approach did not pan out.

I suspect there are niches that you could profitably target. For example eating habits. Or exercise habits. Or studying habits for preparing for an LSAT exam.

In all of these cases the niche allows you to go much deeper with the product to address specific pain points.

You can also go much deeper with your advertising and marketing strategy. You can also reach out to specific podcasts and to specific blogs for marketing exposure.

Re: My app failed: lessons learned

#172

Post by asimjalis »

I don't think the need-to-have versus want-to-have distinction is sufficient to kill a product. If you niche the product down to a very narrow market you might find people who need to change their habits. It could be a matter of life and death for them. For example, someone trying to lose weight, or break an addiction, or study for an exam that has life-changing outcomes for them.

Re: My app failed: lessons learned

#174

Post by bambax »

lambo2991 wrote: post #140
bambax wrote: post #58

Slightly OT: I built myself a journaling app that sends one email every day to remind me to write about that day; I respond to that email and the response gets stored in a db as an entry for that day. The system doesn't try to be clever in any way, it just sends and receives emails; it's proven quite effective. It's been running for about two years now and there are only a couple of days missing, if that. It's very i…

do you mind explaining your stack / tech a bit? would love to whip up something similar for a pet project

Sorry for the late answer.

The db (sqlite) has four tables:

    users
    tasks
    prompts
    items
Users have an email address and own tasks. Tasks are metadata, mainly the frequency (in days) and the time of day when the prompt should be sent. Prompts are a log: each time the system sends a prompt about a task it is logged in the prompts table. And items are entries for tasks: users replies to prompts.

Every x minutes a python script (cron) scans the db to see if a prompt needs to be sent about a task: for each task, when should the prompt be sent, and when was the last prompt sent?

Emails are handled via Mailgun which is used to send prompts and process replies. When a reply is sent, Mailgun creates a POST request to the system with the body of the email.

Each prompt has a unique ID that is present in the respond-to email, so that replies are correctly attached to a prompt (and therefore to a prompt date and a task); it also allows to reply late: each reply is linked to the prompt it replies to, not to the most recent one. It's also possible to send multiple replies to a single prompt.

Web interface for receiving posts from Mailgun and seeing the entries is coded in PHP, but it's extremely light. There is no functionality for updating entries, creating users, etc.: this is done via phpliteadministrator (very useful) or sometimes directly via SQL. There are also triggers in the db to do integrity checks when inserting or updating.

Re: My app failed: lessons learned

#175

Post by somenewaccount1 »

somenewaccount1 wrote: post #37
kylebolt wrote: post #2

So early 2020, I launched Rithm, an app that helps you build habits, reach your goals and stay motivated. The 1st version was a chatbot app, think of it like an accountability coach that you text with to track habits and stay accountable. Rithm landed as the #4 product of the day on ProductHunt, had some good initial traction which proceeded to fall off a cliff. Here are the 8 lessons for a failed version 1, see vide…

Looks like you are still working on learning distribution. I'm typing on a mac, but own an android. i still can't use or try your app :(

aaaaaahhhh.... negative 4 points for actually reading the article/watching the video, and being accurate.

hacker news blowing steve jobs as usual. that's the type of exclusionary behavior i like to see!

Re: My app failed: lessons learned

#176

Post by somenewaccount1 »

somenewaccount1 wrote: post #37
kylebolt wrote: post #2

So early 2020, I launched Rithm, an app that helps you build habits, reach your goals and stay motivated. The 1st version was a chatbot app, think of it like an accountability coach that you text with to track habits and stay accountable. Rithm landed as the #4 product of the day on ProductHunt, had some good initial traction which proceeded to fall off a cliff. Here are the 8 lessons for a failed version 1, see vide…

Looks like you are still working on learning distribution. I'm typing on a mac, but own an android. i still can't use or try your app :(

aaaaaahhhh.... negative 4 points for actually reading the article/watching the video, and being accurate.

hacker news blowing steve jobs as usual. that's the type of exclusionary behavior i like to see!

if anyone payed attention - i even own an mac AND HIS APP DOES NOT WORK ON IT. That barely requires a frigin flag to compile and distribute for, it doesn't even take extra work. But fuck me, right?

Re: My app failed: lessons learned

#177

Post by kylebolt »

dzonga wrote: post #153

one thing, the OP is missing. with consumer oriented stuff, don't make the naive mistake of launching thinking you'll make money. do it as a vanity project. yeah, every now and then shout it out but don't expect it to make money. unless you're vc funded and you're trying to corner a market. b2b is different of course - there exists people willing to pay

Great point!
Post reply on HN