Flatlines, confirmation bias and pretty graphs

Neeraj Singh's Avatar

Neeraj Singh

2025-09-17 11:26:16 UTC

Today I got unsolicited advice from a Redditor. I'll call him MG (Marketing Guru). Here is what MG said.

marketing_cwr1mT8MD.png

MG is talking about NeetoCal metrics. MG is right. If you look at the graph, the first two years it is dead.

bookings_AjTZutvmth.png

In May 2022, NeetoCal had only six meetings booked. NeetoCal had to wait until January 2024 to book the first 1k meetings.

MG is right. If I remove those first two years from the graph, then the graph will look nicer. And I won't be lying. But here is the truth. It won't be the complete truth.

And that's the problem with many founders, VCs, marketing, and sales people. They have a lot of "selection bias".

They glorify startups like Bolt, which generated millions in revenue in four months, and WindSurf, which flipped for billions of dollars. Guess what - that's a clear case of confirmation bias.

Most of the startups end up in the gutter. If a startup makes it to profitability, that's a high-water mark.

So yes Mr. MG I can hide that data to make it look like NeetoCal made it to 40K bookings in a record time. But I won't.

Because I’m not here to impress VCs who get dazzled by those tricks. I’m here to build Neeto products and share our metrics as they are — boring, flatlines, and all. No smoke. No mirrors. Just the real data.

View archive

NeetoPublish Made with NeetoPublish