What fails have you experienced recently?
We have a culture of not talking about failures. It's much more fun to talk about our wins! However, I do believe it can have such a negative and mental impact on us if we constantly expect ourselves to be winning all the time.
Can we break down the barriers by opening up about things that haven't worked?
What have you failed at recently?
#ask-ih
Trending on Indie Hackers
🔥 Roast My Landing Page
I think a lot of the changes I made to the IH forum back in January were a mistake, including: putting posts in modals instead of on their own separate page, defaulting to a feed view instead of the more compact list view, sorting posts chronologically rather than keeping the more popular posts near the top for a while, making post titles optional, etc.
Many of these changes accomplished exactly what I wanted them to, but also resulted in negative second-order effects that I didn't predict at all. For example, the chronological sorting dramatically increased the percentage of posts that get responses, but also managed to make the forum look less lively, since the top posts always had 1 or 2 comments instead of 10+ comments.
All told, I learned a lot from the data and from talking to people about the changes, but it was a pretty big waste of time.
Is that a bad thing? Isn't somebody flailing around and getting no help a bigger problem than getting that 9th comment on a thread?
You'd think so! But when the forum doesn't have any clearly visible popular posts, new visitors react by saying, "This place isn't that popular, might as well not come back." So fewer people sign up, which means there are fewer people commenting (and fewer people getting help), and now it's a vicious cycle. Not something I predicted. Gave me a healthy respect for why companies like Twitter don't tweak stuff that often.
Ah, I hadn't realized short-term sign-ups was a primary goal.
By "vicious cycle" I mean the problem gets worse over the long term, not just the short term.
Ultimately a community is powered by people, so losing more and more people makes a community less and less useful. Ideally you could ask a super niche question and someone would have an answer.
Oh, people were actually quitting over seeing fewer comments per thread? That's really surprising!
I wonder if maybe there's an optimal level. E.g. seeing 10 comments might keep people returning more than seeing 1 comment, while seeing 1,000 would be less appealing than 10.
I think it'd be an exaggeration for me to say people were quitting because of that. Rather, people churn from the forum all the time, and so the rate at which new people who see the forum decide to sign up and engage needs to exceed the churn rate, otherwise it shrinks.
There's definitely some optimal level of visible activity. I don't know what it is, though.
My feeling is that IH needs groups of some sort to spread the attention around. If the homepage is the only hang out spot, then responses percentages are bound to fall as the site grows, because the homepage stays fixed in size as an ever-increasing number of people join and make posts.
My product hunt launch got ... 3 upvotes :(
If you launch and no one notices, then you get to launch again! 😂
A couple of weeks ago I completely reworked my landing page in conversion rate optimization efforts. I added more images, screenshots of the drawing tools, and a lot more copy all within a Pain-Agitate-Fix framework... the end result? My conversions dropped by half.
Oh no... 😒 How was the previous copy structured?
You can see a before/after here: https://imgur.com/a/NpJr7Rx
And since then, I've continued to tweak the landing: https://www.paneljam.com
Not sure where I went wrong... have I made the product more difficult to understand?
Hard to say but I’ve noticed the first version speaks to loneliness and the next to art block. These seem vastly different in the feels approach. It may be that people relate to the theme of loneliness more so than art block. May. Generally speaking, I have noticed that short form pages convert better than long form. Generally speaking. This depends on what you are selling. Why not try something even shorter than first version and get emotional copy that makes them curious to see what that’s all about?
Thanks! I may have to give that a try...
Let us know how it goes! I am very curious for myself.
Losing my job 😬
Even though I knew from the start that I should validate my idea as early as possible, I've still spent at lot of time building the first small product to test the idea. I have a working product, but now I'm at a place where I don't really know my potential customers' real problems, and whether I'm solving them. Also, I'm building for people that often themselves struggle with making money from their products: podcasters, musicians. I think I should have invested more time in market analysis, which is what I'll try to do next. Better late than never.
Do you have a good resource on how to do market analyses?
A couple of weeks ago I completely reworked my landing page in conversion rate optimization efforts. I added more images, screenshots of the drawing tools, and a lot more copy all within a Pain-Agitate-Fix framework... the end result? My conversions dropped by half.
Had to email the investors for my first startup to tell them that we failed. That one hurt. But they were incredibly supportive; the lesson: pick your investors carefully. We did, and when things went to shit, I'm glad we didn't take all of the money we were offered.
My wrists have started getting worse again... so I can be social online via voice recognition but it's put a hard limit on how much I can code (which is at the core of all the content I make).
My mistake was pushing a bit too hard because I was feeling better and making a lot of progress. I was excited and there were a lot of things to do.
That sucks. I hope they'll get better. There's no voice recognition for code?
Yeah, physical limitations suck. There are too many special characters in code, unfortunately.
I have not got any comments to my last post yet. :(
Which? Where?
Here https://www.indiehackers.com/post/a4b9646188
thanks.
Usual trap... built a lot of features for a problem that people probably didn't have. Didn't try and find people that had the problem I was building features to solve :).
Oh well. Lessons learnt.
I'm too focused on my stupid business and my plants are dying
This is so funny. Plants always die on me. My mother keeps buying them for me even though I’ve asked her to stop as it feels kinda like plant cruelty.
But maybe there’s a market to outsource watering of plants for entrepreneurs! Haha.
hahah!
Nice meeting in real life the other day.
Shame we didn't get to talk so much. Till next time :)
I’m seeing your “stupid business” (the Nike piece) everywhere lately (Reddit, HN...)! Congrats! 😀
hahah! thanks andrea
I'm not seeing my plants too much though.
I got the most shade friendly plants for my room and managed to let both get close to dead. I don’t deserve plants :(
hahahah!
Try succulents! I love mine, and they're pretty tough plants. And beautiful, too.
Oh, I've killed succulents too.
In that case, I'm out of ideas. ;)
At Pointr.app project, my biggest failure so far is that I thought creating a rich text editor will be an easy task. Boy, I was wrong. Even with reducing the scope to the bare minimum and trying multiple libraries (and picking one that I think does everything right) I still pretty early in the journey. For the future tasks, I'll timebox "best user experience" and if I'm not able to get it working in that timebox, I'm opting in for a "shitty" default component.
At my day job, I also have tons of failures, where something that we build isn't working as we expect, or customers don't get it, or they use it completely different than what we would imagine. That's part of the job. The great part that we learn from our mistakes and not repeating them again. The greatest part is when after all iterations we finally got something right and it is there to serve customers for the years to come!
This is definitely the thing I would build on top of one set of primitives rather than pulling in multiple libraries. It's totally doable and a problem that's been solved a lot of times, but it's not a trivial task, though.
Most rich text editor features are probably not too hard to implement on your own, without tying yourself to a specific library beyond a base like React or Angular. You might lose as much time fighting the library as you gain from using their abstractions.
'contenteditable' isn't a standard and every browser behaves little differently. The worst part about that is when you do the same exact actions 3 times in the same spot it produces three different htmls. That is a showstopper for my startup, where I need to control diffs.
Slatejs and Prosemirror are giving you a pretty low-level abstractions, but as any library come with their own learning curve. So, I decided to go with that curve instead of the unpredictable world of 'contenteditable', turns the curve is sooo much steeper than you would think. Both libraries aren't helping you with advanced use cases at all :)
PS I now have much more respect to engineering teams of Confluence, Medium, Figma, and Google Docs. They're pulling miracles.
I see, and testing on all platforms/browsers probably isn't feasible for a startup either. I hope it works out, sounds like you're going to have to pull a miracle as well.
Thanks for kind words! I limited the scope to a biteable size that I only need and let the big guys do it for a general audience.
I moved Front Desk to a new domain, and lost almost all of my google rankings. I thought I had everything in place, 301s, updated back links, new content. But it still imploded when google made the switch.
I now believe the piece I missed was creating new back links after putting the redirects in place. Once I got a few back links to https://tryfrontdesk.co (←like this one 😜) the google rankings popped back.
Lost a month of signups to this one, hopefully it will be a long term win 🤷♂️
Lack of focus. Get caught in the excitement of building new features which distracted me from the main initial vision. So I've been neglecting the core offer and ultimately lost... Well may be 6 months or so, to ultimately go back to the initial core offering while those additional features (build stripe integration for instance) are useless at this stage.
Spent a lot of time trying to do everything myself. But eventually realize that a better way to do is to create a process and delegate.
Marketing directly to LinkedIn 1st level contacts. So far only 1% conversion rate. Had been hoping for 5%+. I took my decent sized list, reviewed their LinkedIn Profiles for critical flaws, incompleteness, and missing best practices. Then I did a customized message (using R-project, markdown and mail merge functions) sending a direct message listing the number of errors/omissions, etc. with special offer to my affordable email drip course (including first lesson/outline of course for free) or full service editing. Have tried a few iterations of the DM per each 100 contacts. Disappointed. The process felt a little slimey based on the kind of DMs I often get but I tried to deliver real value and heads-up that fixes were required.