Free Radicals

The chemistry metaphor is apt. A free radical is a molecule with an unpaired electron — highly reactive, bonds easily with other molecules, drives chain reactions. Useful in controlled quantities. Uncontrolled, causes oxidative damage.

Free radical employees work the same way. They don't fit cleanly into org charts. They cross team lines without permission. They challenge assumptions that everyone else treats as settled. They create reactions — some of which are exactly what the company needed, and some of which cost more than they were worth.

The question isn't whether to have them. It's whether your company is at a stage where their reactivity is fuel or fire.

This isn't about brilliant jerks

Netflix drew a hard line on this: no matter how brilliant someone is, if they undermine team dynamics, they go. Reed Hastings has been unambiguous about it — the cost to effective teamwork is too high. That's the right call for the right kind of company at the right stage.

But free radicals aren't necessarily jerks. The distinction is structural, not behavioral. A free radical is someone whose work doesn't sit cleanly inside any defined scope. They're hard to manage because they're hard to categorize. They create value in ways that don't show up on OKRs. They create friction in ways that don't show up on OKRs either.

The brilliant jerk problem is an attitude problem. The free radical problem is a fit problem. They overlap sometimes, but they're not the same thing, and conflating them leads to bad decisions in both directions — keeping the wrong people and losing the right ones.

When they're net positive

The company doesn't yet know what it needs to build. James March's 1991 paper on exploration versus exploitation in organizational learning makes this distinction clearly: organizations need to balance discovering new possibilities against executing what they already know. Early-stage companies have no exploitation to protect. They need maximum exploration, which means they need people who don't respect the existing map — because the map is wrong.

Information is siloed and nobody is authorized to cross the lines. Free radicals tend to span structural holes — gaps between groups that don't talk to each other. They're often the first to see problems that require connecting information across boundaries. A free radical who's had three conversations with customers that the product team hasn't had is carrying signal the company needs.

The strategy is calcifying and the company can't see it. Nokia had process, governance, and accountability. It also had a culture that treated internal skeptics as disloyal. The people who saw the smartphone threat early were structurally incapable of being heard. Free radicals in a healthy, adaptive company are sometimes noise. Free radicals in a calcified company are sometimes the only honest signal left.

There's a turnaround happening. More on this below.

When they turn net negative

The company has figured out what works and needs to do it at scale. Levinthal and March identified the "myopia of learning" — organizations that keep exploring when they should be exploiting fall behind. Once you have a repeatable sales process, a free radical who wants to reinvent it every quarter is costing you real money. The exploration-exploitation tradeoff shifts as the company matures. Protecting what works becomes more valuable than questioning it.

Coordination costs compound. At Series B and beyond, there are enough people that coordination itself is a constraint. When a free radical acts unilaterally across team lines, they're not just doing one thing differently — they're breaking someone else's assumptions about what's stable. At ten people, someone going rogue is annoying. At two hundred people, it's a reliability problem that cascades.

They've misread psychological safety as permission. Amy Edmondson's research on psychological safety shows that high-performing teams have more openness, not less — but safety means you can speak up without punishment, not that you can blow up decisions already made. A free radical who interprets a psychologically safe culture as a license for unilateral action has the concept backwards.

Stage by stage

Seed. Free radicals often are the company. There's no org to disrupt. Everyone is questioning every assumption because they have to. The thing to watch at this stage: even at seed, some things need to stay decided so the team can move. Exhausting every decision is a kind of paralysis.

Series A. You've found something that works. The free radical who helped you find it now needs to be pointed at the next unsolved problem, not turned loose on the solved one. Keep them in discovery roles — new market segments, new product bets, partnership exploration. Get them away from the processes they'll just slow down.

Series B. This is where it gets genuinely hard. The company is adding enough people that it needs coordination structures. Free radicals who were founders' favorites in the early days now create real friction with everyone who joined after them. The right move isn't necessarily to fire them — Tushman and O'Reilly's work on organizational ambidexterity showed that companies can maintain exploration and exploitation simultaneously, but only if they're structurally separated. A skunkworks. A separate team with a different mandate. Not the same team trying to do both at once.

Series C and growth stage. Most free radicals who haven't transitioned into formal roles by now are either already gone or entrenched enough to be a governance problem. The ones worth keeping have either become founders of internal initiatives with real charters and real resources, or have figured out how to operate through influence rather than unilateral action. The ones who haven't made that transition are usually causing damage proportional to how long they've been tolerated.

Pre-IPO. The company is learning to operate like a public company. Everything gets more documented, more auditable, more predictable by necessity. A free radical at this stage who hasn't formalized their role is a liability — not because their instincts are wrong, but because public companies live and die on process predictability. This is usually a departure point. The good departures are mutual and clean. The bad ones are messy and expensive.

Public company. At this scale, free radicals need institutional separation to function at all. Google X, Amazon Lab126, Lockheed's Skunk Works — these aren't accidents. They're the organizational answer to how a large company maintains exploration capacity without letting it undermine execution. The free radicals go there, with a clear mandate, separate P&L, and protection from the core business immune system. They don't run the core business.

Turnaround. This is the exception that clarifies everything else. When a company is in existential crisis, it usually needs someone to do what the organization has been actively preventing. Steve Jobs returning to Apple in 1997 is the obvious example — he cut 70% of Apple's product lines, killed the clone program, and reset the culture in eighteen months. He was, by any structural measure, a free radical given operating authority. The board gave it to him because the company was 90 days from bankruptcy and the internal immune system had already failed. You reach for the free radical when the existing system has demonstrably broken down and the stakes are high enough to absorb the disruption.

The turnaround case is also worth studying because of what comes after. Jobs eventually built Apple into one of the most process-disciplined companies in the world — tight supply chain, constrained product lines, rigorous launch cycles. The free radical who saves the company usually understands that their job, once the turnaround works, is to build the structure that makes free radicals unnecessary.

What to actually do with one

If you're a founder or executive trying to figure out what to do with the free radical on your team right now, start with an honest read on your stage.

Is your constraint exploration or exploitation? If you don't know what to build next, the free radical is an asset. If you can't execute consistently on what you've already figured out, they're probably making it worse.

Separate what they're doing from how they're doing it. A free radical surfacing real problems through destructive methods is a solvable problem — give them a legitimate channel and a real mandate. A free radical surfacing imagined problems is something else entirely.

Don't wait too long to formalize or separate. The longer a free radical operates without a clear scope, the more the organization builds informal dependencies on their unofficial role while also resenting the friction they create. It becomes a mess that's far harder to unwind at Series C than it was at Series A.

And sometimes the honest answer is that they need to leave. Not because they're wrong, but because what they're built for and what the company needs right now are genuinely incompatible. The best free radicals usually know it before anyone has to say it. The worst ones never figure it out.