Why the Accelerator Cohort Is Broken | Mick Kanfi of Earth Collective

What we talked about.

Sunny Ray sits down with Mick Kanfi of Earth Collective to ask whether the accelerator cohort model still serves founders. Kanfi traces Earth Collective's roots to Earth Tech, an Australian company that ran innovation programs and found a readiness gap between startups and investors. A physical challenge held on Richard Branson's private island in Australia showed him that engagement, not investor matching, was the real product. Drawing on his past as a climate tech venture capitalist, he explains why modeling the cohort approach revealed a ten year payback horizon, and why he chose to build an online global platform he calls the Amazon of accelerators. He discusses how fundable is shifting in the age of AI, why impact still takes a back seat while money chases AI, and plans for an AI for good innovation challenge alongside a B2C launch in January. Kanfi also covers a large unnamed pilot with a multi-country network, his 90 day plan, ideal partners, and a five year vision that includes dynamic funds, insurance, nonprofits, and project finance.

A serial founder explains why 98% of accelerator applicants get rejected and how he's building a global, self-serve platform for impact startups.

The questions, and the answers.

What is Earth Collective and what problem are you solving?

We're trying to become a global accelerator for impact startups. We use an open model where founders can self-serve, getting the support they need at the moment they need it, with far more access to support than a traditional accelerator offers. We're delivering it through a global SaaS platform approach.

What did your early programs teach you about investors and startups?

Earth Collective grew out of Earth Tech in Australia, which matched investors and startups. We saw a readiness issue on both sides. Investors said they wanted early-stage deals but wouldn't pull the trigger, and startups fell apart at the second meeting or data room stage. Readiness became the problem to solve.

What happened at the first innovation challenge you ran?

It was called Together in Action. We picked 50 finalists, mostly startups from the Global South, and flew them to Richard Branson's private island in Australia, with around 50 advisors and investors as judges. Engagement and mentorship were amazing, but no investments followed. The lesson was that the engagement itself is the product.

When did you realize the cohort model itself was the problem?

When I did the fund-style math as a former VC. A cohort model means roughly ten years before you might see a return, which is why many accelerators rely on government support. It's also limited by offices, physical space, and fixed start and end dates. I asked what the Amazon of accelerators would look like.

Why are so many applicants rejected, and how does that affect impact startups?

My research shows about 98% of applicants are declined by accelerators. Impact startups get the short end because accelerators look for likely winners who will exit and return money. But the world needs all kinds of solutions, so the lens should be whether it's a good business and product that serves humanity.

What does fundable mean to an investor today?

It's changing, especially with AI. Building a product is now quick, so more attention goes to the founders and whether they can go to market and reach the addressable market they claim. Impact is in a similar spot, because money is following money, not purpose and ethics.

What does your current pilot look like and what comes next?

We're running a big pilot with a multi-country network I can't name yet. Their physical accelerator took 10 of 150 applicants and was expensive to scale. We're getting about 20 applications a day, and it could reach 10,000. Now we monitor, fix bugs, and add features from community feedback.

What is your five year vision for Earth Collective?

Hundreds of thousands of founders on the platform, which pulls investors and advisors in because everyone is there. With all that data, we could spin up dynamic funds around startups that are ready. We could also offer customized insurance and support nonprofits and project funding across the whole assembly line.