Kill the Business That Made You

Winding down a business that still works

By 2022, Prenetics's COVID-testing operation was the company by any operating measure: eight hundred million dollars of revenue over three years, forty thousand tests a day at peak, all inbound-flight testing for Hong Kong, two to three thousand staff, and a NASDAQ listing at a valuation above one billion dollars in May 2022. Danny Yeung shut it down anyway, laying off two thousand people, on a bet against the cash still arriving: "a lot of companies went public during that time and the crazy thing is they're all bankrupt now or not listed anymore... I felt at that time, I don't think testing is going to continue."¹ Peers who read the same demand collapse as a downturn to survive kept their cost base intact waiting for a recovery that never came, a case of Cyclical vs Structural Ambiguity resolved in real time, and the counterfactual is unusually well evidenced: several multi-billion-dollar comparables that made the other call are bankrupt or delisted.

Why it is hard to act on

Nothing about the decision was forced by the numbers at the time it was made. The revenue was still arriving, so there was no loss on the page to point at. The people being let go were the ones who had won the fight, three years of what Yeung calls literally constant intensity, including staff he had hired out of the pandemic's hospitality layoffs. And the identity being liquidated was the company's own: Prenetics was the COVID company in Hong Kong, and everyone flying in had taken its test. His resolution was that the alternative was worse for everyone, "I knew if I didn't let go of them, the company's out," and that releasing people early into a working job market beats releasing them late from a failing one.¹

What made the shutdown survivable

Two things carried through. The balance sheet came out with more than one hundred million dollars in cash against a market capitalization that had fallen to roughly forty to fifty million, and that cash funded the build of IM8, the supplement brand Yeung co-founded with David Beckham in 2023, an application of Capital Allocation Discipline. And the lab, science team, and regulatory posture stayed intact, the same infrastructure that had let the COVID business itself stand up in three weeks in the first place, repurposed a second time.

The counter-case

The opposite counsel comes from a company with a twenty five year record of doing the reverse. 37signals has rewritten Basecamp three times and kept every version running: the first version was pulled from sale in 2010 and sixteen years later still has paying customers producing millions a year in near-pure profit, because a discontinued product carries no development cost and minimal support load.² David Heinemeier Hansson concedes the underlying point rather than denying it, that some products get stuck with their initial cohort and age with it, and resolves the tension by running multiple versions simultaneously rather than sunsetting anything, new product for new customers and a frozen product for the cohort that already solved its problem, a hedge that plausibly requires an owner-operator willing to carry several generations of code that no outside board would approve.

Open question

The story is told by the side that turned out to be right, so Yeung's read on COVID testing reads as wisdom in hindsight; the identical conviction applied to a demand shift that turned out to be cyclical rather than structural would read as panic, and that version of the story does not get told. Listing the company at a billion-dollar-plus valuation driven by COVID revenue in May 2022, while reportedly already believing the demand would end, is a tension the account never quite reconciles either.