Why we fell for the founders building sport's data layer: our investment in Superstat

Sports fans, rejoice! It’s been a stacked few weeks with the World Cup, Wimbledon, The Open, Tour de France, Women’s T20, Comm Games, and the Knicks winning their first NBA title in 53 years. Every one of these moments is getting replayed, broken down and stat-checked, because elite sport generates more data about itself than almost anything else on earth.

And yet, none of that exists for the other 99% of people who actually play sport - including the thousands of kids on suburban courts who'd love to be the next finals MVP. That's the gap Superstat is closing.

This week Blackbird led the Seed round into Superstat, a Melbourne-founded startup turning iPhone footage of amateur basketball games into full box scores, shot charts and highlight reels - no clipboard, no cameraman, no special equipment. We were joined in the round by Startmate and a group of strategic angels.

The simple version of this story is that we invested because the product is great and the market is big - both true - but it's only half the reason we invested. The more interesting story is the team building it.
‍

‍The data nobody's collecting

Amateur sport is one of the strange corners of the world where almost everything is observable but almost nothing gets measured. 

Around 70% of grassroots basketball games are already filmed, usually by a parent with an iPhone rolling on a tripod. The footage already exists, but turning it into usable stats takes someone 2-5 hours with a clipboard, rewinding the same passages of play - and even then, the numbers are often wrong.

Coaches pick starting fives without reliable data, clubs recruit without it, and since the NCAA introduced the NIL rules in 2021, parents have been spending thousands of dollars chasing college opportunities for children whose performance has rarely been measured objectively.

What just changed

For years, solving this problem wasn't commercially viable. Three things have converged in the last 18 months to change this: 

1. Computer vision models are finally good enough to work on messy footage and offer a level of tracking that, until recently, only existed for professional broadcast film. The kind of analysis a Knicks game gets from a dozen broadcast cameras, a suburban Sydney basketball court can now get from one iPhone propped against a fence.
2. The economics now make sense. GPU inference costs have fallen dramatically, making it economically viable to process every grassroots game rather than just elite competition.
3. The input already exists. Parents are already filming games, and coaches aren't being asked to change their workflow or buy new hardware. Superstat simply turns that existing video into structured performance data.

AI can finally give every weekend athlete the kind of performance data that used to be reserved for the pros. That's the opportunity: Superstat is building the data layer for all sports, starting with basketball.

What falling in love looks like 

At the end of our final IC meeting, Cordelia and Kai asked us a simple question.

"How does Blackbird help the founders you back?"

Before I had a chance to answer, everyone on our team came off mute. One after another, they jumped in with different ideas: US expansion, introductions to founders a few steps ahead, company building, hiring. They were all good answers, but I couldn't help but laugh.

Looking back, what we were really doing was falling over ourselves to make sure the Superstat team felt just how excited we were to be part of the journey. It's a rare feeling, but one every investor hopes for - meeting founders you can't wait to spend the next decade working alongside.

I first met Cordelia eleven months earlier, on a walk around Barkley Gardens in Melbourne. Superstat was three months old and was building computer vision models for AFL.

Looking back, what strikes me isn't how much the product has changed. It's how much bigger the founders now believe it can become.

Back then, the ambition was building software for AFL clubs. Today, they're building the data layer for sport.

Kai and Cordelia have known each other since primary school. Sam joined them at 16 to build their first company, Trainstop, which became Australia's largest sports recruiting platform. Trainstop is where they kept running into the same problem: clubs wanted to recruit using data that simply didn't exist. Superstat is the company they built to fix it.

But ambition alone isn't enough. What convinced us was the pace at which they kept raising it.

Superstat launched in March 2026. Through Startmate, the company compressed months of learning into weeks: by week four they'd dropped selling to clubs in favour of parents, by week six they'd scrapped subscriptions, by week nine they'd launched in the US, and by week ten they'd committed to relocating there.

Every success was met with a more ambitious goal. Watching a team compress that much learning into ten weeks tells you more about them as founders than any pitch deck could. 

Sixteen weeks after launch, Superstat had processed 2,363 games and grown to 5,135 monthly active users.

None of them have taken a salary. Kai sold his car to fund model development, Sam dropped out of university, and Cordelia moved back home and gave up her income. They've spent the better part of a decade chasing the same problem, and every sign of progress has only made them think bigger. 

We're proud to back Kai, Cordelia and Sam as they build the data layer for sport.

Maddy Guest, on behalf of Blackbird

‍