Ekcho

Two days ago I posted about launching Ekcho — a pseudonymous voice broadcasting app. The response here was genuinely warm and I want to follow up properly because several of you asked questions worth answering honestly.

First — thank you.

The IndieHackers community gave me more thoughtful feedback in 24 hours than I got in weeks of building alone. Particular thanks to everyone who flagged the moderation challenge, you're right, and I'll come back to that.

To the question: what kind of users are signing up?

Honest answer, it's too early to pattern match with only 10 signups. But what I can say is that the people who have signed up aren't who I expected. I assumed early adopters would be creators, people who already have something to say publicly and want to say it without their name attached.

What I'm actually seeing is more like listeners first. People who found the concept interesting, signed up, explored the feed, and haven't yet recorded anything. Which tells me the barrier to the first broadcast is higher than the barrier to signing up. That's my next problem to solve.

To the question: do people behave differently when anonymous?

This is the question I find most fascinating and the honest answer is, I believe yes, fundamentally, but I don't have enough data yet to prove it on Ekcho specifically.

What I do know from building this is that in testing, people said things in voice recordings they told me they'd never post publicly. Not controversial things, just honest ones. The absence of identity seems to lower a specific kind of self-censorship. Not the moral kind. The social performance kind.

The thing I didn't anticipate is that voice anonymity feels different from text anonymity. Text anonymity has a reputation for cruelty. Voice anonymity seems to produce something closer to vulnerability. You can hear when someone means what they're saying. That changes the dynamic entirely.

To the moderation point:

Several of you flagged this and you're right to. The moderation pipeline I built runs every broadcast through CSAM detection, then OpenAI Whisper transcription, then OpenAI Moderation API with custom thresholds. Crisis language triggers immediate admin notification. I’m refining this and will see an update on this in the coming weeks and months as user base grows.

Is it perfect? No. Will it be tested heavily? Almost certainly yes. But I made a deliberate decision to build safety infrastructure architecture plan before user acquisition infrastructure. Although it’s still in testing phase. I'd rather launch slow and safe than fast and harmful.

To the positioning feedback:

One commenter said it perfectly, it's not just "voice broadcasting," it's "a place to speak freely without identity pressure." That's sharper than anything I've written and I'm going to use it. Thank you.

On the demo video point, completely agree. A 15-30 second video showing the journey from recording to publishing to listeners would make the concept click instantly for people who don't get it from text alone. Working on this now.

The honest numbers update:

• 28,000 impressions

• 1,400 clicks

• 10 signups

• 0.71% conversion rate

CTR is strong at 5%. Conversion from click to signup needs work. The landing page has been rebuilt this week and I'm watching whether that moves the needle.

What I'm doing this week:

• Building the demo video

• Fixing the first broadcast friction, making it easier and less scary to record your first thing

To the person who mentioned anonymous accounts on X and Reddit:

You're exactly right. There's a whole population of people who have built anonymous text followings who might find anonymous voice genuinely interesting. That's an audience I haven't gone after yet and probably should.

If you're reading this and you have an anonymous online presence somewhere, I'd love to talk to you. What would make you try anonymous voice? What would stop you? Email me at info@ekcho.net or comment below.

One last thing:

Ekcho is three days old. The numbers are small. But the comments here have given me more clarity about what I'm building and who it's for than months of building in isolation did.

That's the real value of posting here. Thank you for that.

ekcho.net — free to try, no real name needed.

— Kismis03

Hey IH community,

Launched Ekcho this week after months of building solo. Sharing the journey here.

The idea

I kept noticing that people edit themselves online because their identity is always visible. I wanted to build something where the voice earned its own reputation completely separate from the person behind it.

Ekcho is a pseudonymous voice broadcasting app. Pick a voice name. Record. Publish. No real name, no photo, no identity on the line.

The build

Solo founder. Built on Next.js 14 with Supabase for database, auth, storage and realtime. OpenAI Whisper for audio transcription and moderation. Lemon Squeezy for payments since it works from India and handles VAT automatically. Deployed on Vercel.

Took several months from idea to launch working nights and weekends.

The features that took longest

The moderation pipeline was the hardest part. Every audio file goes through CSAM detection, Whisper transcription, then OpenAI Moderation with custom thresholds. Crisis language triggers immediate admin notification. Getting this right before launch was non-negotiable.

The listener room feature a live chat space that opens automatically when a broadcast hits 50 unique listeners — was also complex to build with Supabase Realtime.

Where I am now

Live at ekcho.net. Early signups coming in. Running Google Ads. Listed on PeerPush and Product Hunt.

What I'm figuring out

Conversion rate from ad clicks to signups is lower than I'd like. Working on the landing page and onboarding flow. Would love thoughts from anyone who's been through this.

Happy to answer anything about the build, the stack, or the journey.

ekcho.net