The Idea

In early 2025, Pierre-Eliott Lallemant had a 2,000-person waitlist and nothing else.

He and Romàn Czerny had just sold CoCo AI, a WhatsApp marketing tool for e-commerce brands they had bootstrapped to €500K ARR on cold outreach alone. No ads. No Product Hunt launch. Just a repeatable playbook and an American fund willing to pay for it.

So the second act should have been easy. They built an AI sales assistant that auto-filled CRMs and took meeting notes, opened a waitlist, watched 2,000 people sign up, and then launched to zero sales. Not slow sales. Zero.

The post-mortem was one line long: interest is not intent. Two thousand people were curious about the product. None of them were in pain about it.

So they went back to the one thing that had actually hurt during CoCo. Not closing deals, but finding people worth closing. Every day the two of them had burned hours in LinkedIn tabs and stale lead databases, hunting for anyone who might be in the market that week.

That was a real wound, and they knew where it was. So they built the thing that found the buyers instead of the thing that filed the paperwork. This was the beginning of GojiberryAI.

The Execution

• Early 2025: Lallemant and Czerny exit CoCo AI, bootstrapped from zero to €500K ARR (roughly $50K MRR) on cold outreach with no paid acquisition, and sell to a US fund. Third co-founder Dylan Teixeira sells Edusign, a paperless attendance SaaS he had grown to multi-million ARR, the same year.
• Mid 2025: Version one of Gojiberry, an AI sales assistant for CRM autofill and meeting notes, launches to a 2,000-person waitlist and makes zero sales. The team kills it and rebuilds around intent detection: job changes, competitor engagement, funding announcements, hiring spikes.
• Late 2025: The pivot lands. Gojiberry goes from zero to 100+ paying customers in 60 days and $33K in revenue inside four months, sold the same way CoCo was sold, by running outbound on themselves.
• November 2025: First Product Hunt appearance as Gojiberry AI, pitched as "Lovable for GTM teams".
• February 2026: Gojiberry crosses $1M ARR in 7 months and 21 days, fully bootstrapped, no outside capital, no sales team.
• March 2026: Second Product Hunt launch pulls 412 upvotes. By the time the team applies to Y Combinator they are already at roughly $2.8M ARR, applying as an accelerant rather than a lifeline.
• Spring 2026: Accepted into YC's Spring 2026 batch and relocated from Paris to San Francisco, with YC's standard $500K as the first outside money the company has ever taken. Team size: 7.
• June 2026: The YC launch post reports $0 to $1.4M ARR in 9 months, $112K MRR, 44% month-on-month growth and 1,000+ paying customers. The growth engine is the product itself: 5 LinkedIn accounts running 30 connection requests and 30 DMs a day, plus 295,000 cold emails in 90 days across 41 domains for about $600 a month in infrastructure.
• Today: Gojiberry reports $2.5M ARR in 10 months, 30% month-on-month growth, profitable, and 2,000+ customers, with Lallemant claiming $3.5M ARR at the 11-month mark and a stated target of $10M.

The lesson?

A 2,000-person waitlist is not traction. It is a list of people who liked the idea of your product enough to type an email address, which costs them nothing and tells you nothing. Gojiberry only worked when the founders stopped guessing at a market and rebuilt around the specific hour of the specific day that had annoyed them personally for two years. Then they pointed the product at itself and let it prove the thesis on its own pipeline. Interest is free. Intent is the only signal that pays.