Article
Hero mode is a defect
Hero mode is when a company runs only because one person keeps rescuing it. It is a process defect: OKRs and PDCA remove the need for the rescue.
You know the person. The message arrives at nine in the evening, they fix it, and on Monday someone thanks them in the meeting.
Everyone in the room reads that as commitment. I read it as a symptom.
I call it hero mode: an organisation that runs only because one person keeps rescuing it, rather than because a system does. It isn’t a sign of a strong team. It’s a defect in the process, and the test is simple. If the same rescue happens twice, the process is the problem.
I learned where companies slip into it as COO across The Search Initiative, Search Logistics and SUSO Digital between 2019 and 2022. The second half of this piece is about how we run against it at NON.agency, my company. That part is self-promotion, so read it as a case, not neutral advice.
Why teams drift into hero mode
Teams drift into hero mode because three things compound: the rescuer is rewarded, the fix is never written down, and volume grows faster than the workflows.
The reward comes first. Nobody gets a bonus for the incident that never happened, but everyone remembers the person who stayed late to fix one. So the organisation pays for rescues and quietly stops paying for the work that would prevent them.
Then the fix disappears. The rescue works, the pressure drops, and writing it down feels like a luxury. A task done well once by someone who won’t document it is a one-off. Hero mode is one-offs, repeated.
Last, volume. A workflow that holds at one size cracks at the next, and the person who knows where it cracks gets pulled in each time. At Search Logistics I turned that into one line on my LinkedIn profile: “I turned manual, ‘one-off’ tasks into a repeatable system.” That sentence is the whole fix, and it’s also the whole job.
What hero mode costs
Hero mode costs you three things: a single point of failure, quality that depends on one person’s week, and no way to hire into it.
The single point of failure is obvious and still ignored. If the whole operation needs one person to be reachable, a holiday, an illness or a resignation is an outage.
Quality is less obvious. When one person patches everything, the output reflects their energy that week. The client can’t see it, but the variance is there.
The hiring cost is the one that stings. You can’t train a newcomer on a process that lives in someone’s head. My LinkedIn profile describes my time at Search Logistics as “hiring the right people, opening new offices, and building processes that actually work.” The order matters: the process is what makes the first two possible.
Not every rescue is a flaw. When a company starts with a five-thousand-pound debt, as Neadoo Digital did in 2011, the founders rescue things because nobody else can. That’s survival. The defect starts when it outlives the need. I saw the manual version at scale at Husky Hamster, the outreach agency I co-founded: the sheer manual effort and human coordination it took to build authority in several languages is what pushed me to look for a better way.
How we run against it at NON.agency
At NON.agency we replace the rescuer with a cycle: OKRs and PDCA instead of gut feel.
Every quarter there’s one WIG, a goal that, if missed, makes the rest irrelevant. It’s broken into objectives and key results. A key result is a number you can check, so it gets discussed less and verified more. Each one has an owner, a target and a status, and a missed one is marked missed. Nobody quietly rescues it in the background.
The other lever is structure. I removed the account management layer: clients work directly with the people building their strategy. Every relay point between client and builder is a place where someone has to translate, chase and rescue. Fewer relays, fewer rescues.
The companion question is whether to hire or build. This is the other half: even a lean team needs a system, or it just becomes a smaller hero mode. More on the company is on the NON.agency project page.
PDCA in four steps
PDCA is a loop with four steps, and the third one is where most teams cheat.
- Plan. Write the key result down: the number, the owner, the date.
- Do. Run the work as planned.
- Check. Compare what happened with the key result, not with how it felt.
- Act. Close the gap before the next cycle: change the process, then write the change down.
The last step is the anti-hero step. A rescue fixes the incident. Act fixes the process that produced it. That’s the difference between a team that firefights and one that gets quieter each quarter.
Hero mode compared with a system-run team
The two models differ on four things you can see without a consultant: who fixes the miss, where the fix is stored, what happens when the key person is away, and how a new hire gets up to speed.
| Hero mode | System-run team | |
|---|---|---|
| Who fixes a miss | The same person, again | Whoever owns the key result, using the written process |
| Where the fix lives | In one person’s head | In the workflow, after the Act step |
| When the key person is away | The operation stalls | The cycle continues |
| Onboarding a new hire | Shadow the hero and absorb it | Learn the documented workflow, then improve it |
None of this is complicated. It needs a habit: write the fix down, and mark the miss as a miss.
The bottom line
Hero mode isn’t devotion. It’s a process that hasn’t been built yet, and it gets worse the faster you grow.
For you, the practical move is small. Take the last rescue your team celebrated and ask what would have made it unnecessary. Then write that down before the next one.
What I can’t tell you is how well this holds as a company grows far beyond its current size. We’ve run NON.agency this way since 2022, and I won’t claim more than that.
The same logic, a repeatable process over a lucky win, applies to getting named in AI answers. If you want more on how I think about systems, the operations and scaling section has it, my podcast Globalne Horyzonty is where I do the research, and the about page has the background.
Questions people ask
What is hero mode in a company?
Hero mode is when an organisation runs only because one person keeps rescuing it, rather than because a system does. It looks like commitment from the outside. Operationally, it is a defect in the process.
How do you know a team is in hero mode?
Look for the same rescue happening twice. If a missed deadline, a broken handover or a late fix is solved by the same person each time, and the fix isn't written down, the process is the problem.
What is a WIG?
A WIG, or wildly important goal, is the one goal per quarter that makes the rest irrelevant if it's missed. At NON.agency it's broken into objectives and key results that can be checked.
How does PDCA stop hero mode?
PDCA forces every miss back into the process: plan against a key result, do the work, check the result, then act on the gap before the next cycle. The fix gets written into the system instead of living in one person's head.
