Five months into building product analytics for conversational AI. Started by targeting vibe coding tools like Lovable but realized most of them don't care about user experience yet. With monthly churn over 50%, they focus on acquisition, not retention.
Now shifting to established SaaS companies adding AI assistants to their existing products. Some of them literally have people reading chats full time, so they actually value the experience.
Building https://lenzy.ai - 2 paid customers, 2 pilots, looking for more and figuring out positioning.
The pivot from vibe coding tools to established SaaS makes sense. Those early AI tools are in pure growth mode - they'll worry about retention once they hit some ceiling.
Established companies adding AI assistants is interesting because they already have baseline metrics to compare against. They know what good user experience looks like in their domain, so when the AI chat experience is terrible, it's obvious.
What's the biggest gap you're seeing between what these companies think they need to measure vs what actually matters for AI chat quality?
You don’t need to identify your users in any way so privacy shouldn’t be a concern in most cases. Lenzy is also super-easy to self-host - just reach out to us at hello@lenzy.ai
Building https://lenzy.ai - helping conversational AI products (think Lovable or Cursor) reduce churn and prioritize product improvements by analyzing their user's chats.
I started about 4 months ago, made my 2 paying customers happy. Now trying to onboard more and more companies!
Building https://lenzy.ai - helping products built around chat with AI (think Lovable or Cursor) reduce churn and prioritize product improvements by analyzing their user's chats.
I started about 3 months ago, focusing on making my 2 early adopters happy. One of them is ready to start paying soon!
Building https://lenzy.ai - helping products built around chat with AI (think Lovable or Cursor) reduce churn and prioritize product improvements by analyzing their user's chats.
I started about 2 months ago, found 2 early adopters and focusing on making them really happy.
Ladas and skoda’s where reasonably common in the UK in the late 80s/early 90s, I always had a bit of a soft spot for them, seeing Skodas resurgence after VW took over was cool as well, Skoda went from a laughing stock to winning car of the year pretty quickly and now people generally like the brand.
There is a Yugo parked near a brewery in Queens NYC that always blows my mind. It's still in decent shape for its age and must be driven regularly as it's always in a different parking spot.
Or the one that made it into an Australian TV advert. Guy walks into a service station.
“Got a windscreen wiper for my Lada?”
“Yeah, mate. Sounds like a good deal to me.”
The classic claim was that the Trabi was made out of cardboard.
Of course, that's a myth: the Trabi was actually made out of cheap plastic.
The Trabant was actually a decent modern car when it debuted in 1957. The problem is that they produced it until 1991, when it was far from modern.
I was born in Zwickau, where the Trabant was produced. It's no accident that they picked Zwickau for the production, because that's where Audi's predecessor company (Horch) had made their cars before.
(Going on tangent: Audi is Latin for 'listen', and Horch is German for 'listen'.)
> the Trabi was actually made out of cheap plastic
Today it would be called an 'advanced composite material', e.g. it's closer to fiberglass than plastic and used recycled materials: https://en.wikipedia.org/wiki/Duroplast
The Trabi was made of duroplast. Sussita, the only Israeli car (similar vintage) used fiberglass (only slightly better). I guess both had the advantage of being lightweight (and cheap).
The Trabant was a normal car if you look at what the rest of Europe was driving at the time. In fact the DDR did not economically collapse until the late 1970s.
Ofcourse the real problem was they could not actually make enough of them. You could go to a Citroen dealer and pick up a 2CV the same day.
My weekend project, FormZero, a free form backend that is easier to self host than to sign up for a paid service, just got an update. Users can now receive email notifications when people submit their forms - wait lists, newsletter signups, surveys.
My first idea was to ask users to set up a free Resend account and use their API key to send emails. While free, this requires users to at least own a domain and definitely goes against my claim for one-click self hosting.
Then I realized that every user already has their personal email address. If only FormZero could send emails from it in a secure way.
SMTP to the rescue - it's the protocol your email client (Apple/Notion/Outlook) uses to send mail from your email address. The fact that it's a standard protocol allows users to connect to any email provider - Gmail, Proton, Outlook, iCloud or even Resend - just bring your sweet SMTP password with you.
This makes FormZero one more step closer to matching paid services in functionality. Next weekend: Captcha and spam protection.
Thanks for sharing. I genuinely enjoy talking to people, figuring out and solving their problems. That's definitely what I quit my job for. But I still consider money a good proxy for how good you are at helping people
What fascinates me about this approach is how it lowers the barrier for self-hosting. I'm curious if there are other projects using "Deploy to Cloudflare" buttons to make self-hosting accessible? Or similar approaches with other platforms (Vercel, Railway, etc.)?
Now shifting to established SaaS companies adding AI assistants to their existing products. Some of them literally have people reading chats full time, so they actually value the experience.
Building https://lenzy.ai - 2 paid customers, 2 pilots, looking for more and figuring out positioning.
reply