I’ve killed more products than most people launch

Do not index

People on the internet have a weird habit.

They see someone’s success - and immediately doubt that.

And I think it’s not their fault, as everyone just talks about their wins. No one talks about their failures.

Everyone sees Revid.ai at $600k MRR. Outrank.so at $300k MRR. SuperX.so at $25k MRR.

Nobody sees the products I worked on that failed.

7 products I killed. Some made money. All taught me something.

Here's what each failure taught me 👇

1 - ProductLift:

The idea: Search engine for communities where you can post when you launch.

Got a few sales immediately. Then *crickets*.

I completely missed something - people only launch once. Maybe twice if they pivot.

You can't build a recurring business on something people need once.

The lesson: Recurring use beats one time value.

This is why I'm obsessed with tools people use daily now. Outrank's backlink checker gets used every week. Revid's video editor gets used for every piece of content. SuperX's aAI features are used daily by users.

If your product solves a problem people have once, you don't have a SaaS.

Recurring revenue = recurring value.

2 - Tweet Butler:

The idea: Tweet generation with GPT-3.

Built it. Launched it.

But AI outputs were terrible.

The problem I didn't see: GPT-3 in 2021 wasn't good enough for this. The tweets sounded robotic. Generic. Nobody wanted to post them.

The lesson: Timing is everything.

You can have the best ideas, but launching them at the wrong time will ensure that you fail.

But here's the interesting part: I killed Tweet Butler in 2021. By 2022, GPT-4 was good enough that Tweet Hunter's AI feature became one of the most used parts of the product (I later sold that product for $8m)

Timing can sometimes be your biggest play.

Some ideas are just 2 years too early. That doesn't make them bad ideas. It makes them bad ideas for right now.

3 - Tweet Butler Digest:

The idea: Twitter actionable analytics powered by AI, delivered daily via email.

Got to $300 MRR quickly. Then stalled. People don't want analytics in their email. They want a full dashboard they can dive into.

The lesson: A feature pretending to be a product will always feel incomplete.

This one hurt because $300 MRR felt like validation. It wasn't. It was people trying it because it was cheap.

Products need to be complete solutions, not clever side features.

4 - ConvertRec:

The idea: GIF and short video creation for product launches.

I didn’t realise I accidentally made an agency.

We got clients. Then realized we were doing custom work for each one.

Again, AI wasn't good enough to automate this. Every video needed manual tweaking.

The lesson: If you have to do custom work for each customer, you don't have a product.

I don’t like agencies. Nothing against them, but I hate running them. You trade time for money. You can't scale without hiring. Every client needs hand holding.

The moment ConvertRec started feeling like client work, I knew it was dead.

If a product can't work without me, it's not a product, it’s a service.

5 - GeoLeads:

The idea: Geo based lead scraping from Google Maps.

Got to $1k MRR. Then realized our best customers were enterprise clients who needed daily communication and a lot of calls (which I hate).

The lesson: Customer type matters more than revenue.

$1k MRR from the wrong customers is worse than $0 MRR.

So I killed the product because the customers it attracted weren't the customers I wanted to serve.

This changed how I think about positioning. My products are self-service or nothing.

6 - AnimStats:

The idea: Create short animations from stats.

This one is very recent, possible that alot of you also know about this.

We quickly grew to $1k MRR. Then stagnated because churn was brutal. People create stat animations once per quarter…

…maybe

The lesson: Monthly subscriptions need high frequency use cases.

AnimStats had the opposite problem of ProductLift. ProductLift was used once ever. AnimStats was used once per quarter.

That's not enough to justify a monthly subscription.

Users would subscribe, create 3 animations, cancel. Repeat every few months when they needed it again.

High churn kills growth. You're running on a treadmill, replacing churned users instead of growing.

Now I only build products with daily or weekly use cases.

7 - Daily Nuggets:

The idea: Personalized daily news, tailored to what you're into.

Traction was low and costs were too high to make it work.

Personalized news requires expensive infrastructure. AI costs, data costs, and delivery costs.

Also, I realized people don't only want news; they want the take of a person they consider an expert.

That's what we call authority, and AI doesn't have authority

The lesson: Unit economics matter from day one.

This one stings because OpenAI basically built this exact product later and put it in ChatGPT.

Same idea exactly.

But here's what I learned: being right about an idea doesn't matter if you can't make the economics work.

Sometimes you're right, but you're too small to execute.

What actually matters

None of these were failures in my eyes.

They were datapoints.

Every "failed" product made my next product better.

Revid.ai works because I learned recurring use matters.

Outrank.so works because I learned to avoid enterprise clients.

SuperX.so works because I learned to wait for the technology to be ready.

Fast failures are cheap education.

I killed these products in weeks or months, not years. Each one cost me less than $5k and a few weeks of time.

Compare that to founders who spend 2 years building something nobody wants.

Launch fast. Kill fast. Learn fast.

One more thing…

Last week someone on Twitter said: "If I was making $1M/month, I wouldn't be writing newsletters."

But I still show up here every week:

Why?

Because of the trust it builds and the support you show.

I get DMs from people telling me:

Someone who closed their first SaaS client using the DM templates

A founder who killed a bad idea after reading about validation (saved 6 months)

Three people who started building their first product because they saw it's possible

That's not something you can automate.

So here's my ask:

Reply to this email and tell me one thing you got from this newsletter, these emails make my day 👇🏼

I read every reply and often write back.

That's why I still love to do the newsletter.

Tweet of the week

banger. everyone should follow.

That's it for this week.

Until next week,

Keep building

Tibo 💻

P.S. If you're working on something right now that feels like it's not working, ask yourself: Is this a bad idea, or is it bad timing? Sometimes the answer is "kill it." Sometimes it's "wait 6 months and try again." Knowing the difference is the entire game.