Connect with the Product Hunt community

Giving users a 40% discount for paying annually felt amazing when we needed upfront cash to fund a month of development. But looking at our servers and support load months later, those deeply discounted annual accounts locked us into low-margin support commitments while our costs kept rising.

at what stage of growth did you stop offering steep annual discounts and start charging what the product was actually worth?

We simplified a setup flow that used to take about ten minutes of configuration. Fewer options, fewer decisions, faster to the first result. A group of experienced users pushed back hard. They said the product now felt like it didn't trust them that the old settings were how they'd made it fit their actual job and that "simple" meant "generic." We'd optimized for the person starting and accidentally made the product worse for the person staying. The simplification was real but so was their loss. How do you decide when to simplify for new users without stripping away the control that keeps the experienced ones around?

Someone used our product for about five minutes and wrote a short, calm message explaining what they were trying to do and why our approach didn't fit. No feature request, no complaint, no suggestion. Just a clear "not a match." It did more for our direction than a year of "looks cool" and "nice idea." It made us realize we'd been sanding the product down to accommodate people who were never going to stay, while the people who actually needed it got a diluted version. The rejection was honest, and honesty is rarer than praise. How do you tell the difference between a user who's wrong for the product and a user who's telling you the product is wrong?

We used to let people use the core function without an account at all. Turn up, get a result, leave. This week we put a free account in front of it. Same allowance, same features, no card but you sign up before you get anything. The reasons were boring and real: we couldn't attribute anything to anyone couldn't tell a returning person from a new one, and an anonymous endpoint that costs money per call was an abuse problem we'd have to solve eventually. What we didn't think through is what to say about it. The people affected are almost by definition people we can't contact. They never gave us an email. That was the whole point of the thing we removed. Every draft announcement reads either defensive or like we're dressing it up as a feature. How do you communicate a change that is genuinely worse for some users without it sounding like spin?

There is a massive trap in expecting an AI to reliably output matching JSON schemas or proper frontmatter without breaking formatting under the hood.

We finally cut out model generated paths and dates entirely, handing those chores over to a simple local hook that feeds exact parameters before every request. The model only focuses on narrative text now, and validation errors dropped to zero.

How do you handle strict data formatting in your agent workflows without constantly dealing with broken syntax?

There is a massive trap when you first start vibe coding where you end up overbuilding features just because the AI makes it ridiculously fast to spin them up.

You spend days adding things nobody asked for only to realize the real bottleneck is maintenance API costs and actually getting people to care.

Shipping the app to the App Store turned out to be the easy part compared to distribution. If you ve built something using AI tools recently what is one thing you wish someone had warned you about before you started?

In most teams I talk to, the content exists: videos, docs, a Notion page. The problem is that half of it never gets finished, and nobody notices until something goes wrong. What s worked for you? Shorter formats, deadlines, nudges from managers, something else?

We hit a weird breaking point last month when I looked at our support logs and realized a single free tier user had opened over 30 tickets asking for custom API work and endless Zoom calls.

At the end of it all they flat out said they d never upgrade because the free plan already gave them everything they wanted. It felt like running an unpaid consulting agency for someone who had zero intention of paying.

How do you draw a hard line on support for free accounts without looking like a terrible company?

I had my agent hallucinate the wrong day of the week three times in a row simply because I let it calculate timestamps on its own instead of using a basic system script.

Moving all structured data generation like JSON fields, file paths, and dates away from the model and into deterministic Python scripts completely fixed our pipeline stability.

LLMs are great for writing plain text and copy, but terrible at programmatic math. Do you still let your agents handle raw config values, or do you lock those down with scripts?

One thing I keep struggling with while building products is figuring out which features are actually worth spending time on.

I sometimes come up with ideas that sound great, but then I wonder whether anyone would really use them. I also don't want to spend days building something that people don't actually need.

I'm trying to get better at understanding user needs before jumping straight into coding.

For those who've built and launched products, how do you usually decide what to work on next? Do you talk to users, look at feedback, or just build your ideas and see what happens?

We brought in a brilliant senior dev from a big tech company early on and it turned out to be a complete mismatch. They were so used to massive teams, automated CI/CD pipelines and fully specced Jira tickets that facing our messy duct taped codebase completely paralyzed them.

They wanted enterprise grade architecture when all we needed was a quick MVP shipped by Friday. For founders who survived their first engineering hire how do you screen out people who need massive support systems before they join?

There is a massive difference between a developer who can write clean algorithms in an interview and one who actually knows how to build something from absolute zero with no product manager or QA team around. Our biggest hiring mistake was optimizing for resume prestige instead of looking for someone comfortable with ambiguity and broken infrastructure.

If you're building a tiny early stage team right now what kind of practical test do you give candidates to see if they can handle startup chaos?

I'm at a decision point in my tech career: one option is a stable corporate role that pays well and looks good on a CV. The other is an early-stage startup that's uncertain, pays less, but I actually care about it. I know everyone faces this. I'm not looking for advice, I'm looking for outcomes.

The part that makes it hard is that both options sound reasonable when you explain them to someone else, and only one of them feels right when you're alone with it.

So for those who chose the safe tech role: did it actually feel safe after you took it, or did the regret come later? And for those who chose the startup: did it work or did you have to go back? And the harder follow up: would you make the same choice again knowing what you know now?

I used to waste an entire afternoon every single month jumping across multiple dashboards just to download billing statements. I finally wrote a background job to handle it all automatically. It s completely unglamorous and breaks the second a website updates its login button, but it gets the job done without human panic. Did you build any low-key scripts lately that saved you from a boring monthly routine?

There s a dangerous trap in being overly helpful to non paying users in the early days. We spent thousands of dollars worth of team hours solving custom problems for a free user who never intended to pay a single dollar.

Giving high touch support to build goodwill sounds great until your team is exhausted and your burn rate spikes.

How do you set strict support boundaries on free plans without pissing off potential leads who might actually convert later?

For months the single most requested thing was a chat style interface. Users said they wanted to talk to the product instead of clicking through it. We built it, shipped it and watched core metrics slide within two weeks. The people who asked for it used it a few times and drifted. The format had changed how they related to the tool from reflecting on what they were doing to just asking for an answer. We pulled it and replaced it with something guided with pauses built in. That version worked better than both. But we only found it because we questioned why the request existed, not just what it was. How do you separate "users want this" from "users will actually use this" before you build it?

Using AI lets us ship updates and write code at ten times the normal speed, but it also makes it dangerous to chase useless ideas.

I caught our team spending a whole week polishing a complex feature just because ChatGPT made it effortless to write only to realize zero users actually wanted it. Shipping fast feels great but it easily masks the fact that you aren't talking to real customers.

How do you stop AI from tricking you into feeling productive while building things nobody asked for?

AI completely stripped away our technical execution bottlenecks but it didn't speed up customer acquisition by a single day.

You can generate landing pages, refine your roadmap, and push code in minutes but getting a prospect to open an email or pull out a credit card still takes the exact same effort as before.

We had to train ourselves to stop measuring daily progress by how much work got finished and focus strictly on revenue and user feedback instead.

If you're building with AI right now, how do you track whether your speed is actually translating to business growth?

We had a bug for months that quietly broke one specific workflow and the users who hit it had built a workaround around it a second tab, a manual step, a small ritual. When we finally fixed it those users didn't feel relief. They felt like the ground had moved. One of them told me they'd spent two weeks learning the workaround and now it was wasted. The fix was correct but it arrived without warning, and it invalidated effort they'd put in. We assumed "fixed" would feel like "better" but for the people who'd adapted, it felt like starting over. How do you ship a fix to something people have already worked around without making them feel like their effort didn't matter?

I spent an hour reviewing a diff today where the agent decided my database schema needed a modern upgrade

even though I just wanted to add a simple validation check.

It s exhausting having to babysit the diff just to make sure it didn't touch background logic.

When you're early post-launch, everyone has an opinion on what your product is missing. It's so easy to get pulled in five different directions trying to please the loudest voices.

I've caught myself almost rebuilding the roadmap just because two people asked for the same feature on the same day.

How do you filter out the noise and figure out which feature requests will actually move the needle for retention?

I built a couple of web-apps that barely have or have none users, and because of that I'm really stuck right now. I can't get any feedbacks on how the web-app is or how could I improve it. Could someone share a working advice?

I was the youngest person on a project once. The least experienced. The one who did not know the acronyms or the history or the unwritten rules.

I spent the first few weeks trying to pretend I knew more than I did. Nodding along. Smiling at things I did not understand. Googling under the table.

I used to waste an entire afternoon every month logging into a dozen different dashboard portals just to download PDF bills. I finally wrote a messy background script to pull everything automatically into one folder. It's completely unglamorous and breaks whenever a website updates its login button but it's totally worth it. Did you build any low-key automations lately that saved your sanity?

When you live inside your own product every single day, you forget what it's like to see it with fresh eyes. Every button, label, and navigation menu feels completely obvious to you.

Then we watched session recordings of our first few real signups. It was painful. Users were getting completely stuck on things we assumed were intuitive, dropping off before they even touched our primary feature.

For builders and indie hackers: what methods or tools finally opened your eyes to your own onboarding blind spots?