Strategy Salons (Nerd Clubs)
Seed a small, opt-in, yes-and idea group that spins off strategy insights as a side effect.
- Difficulty
- Advanced
- Time to result
- ~months to results
- Steps
- 6
- Confidence
- 95%
When a complex, ambiguous, cross-team problem cannot be solved by executive reviews (which produce expensive pageantry that obscures more than it clarifies), Komoroske instead seeds a small voluntary group — a 'nerd club' — with explicit yes-and norms, a slow drip of diverse perspectives, and no decision-making authority. Because nothing is at stake, people build only on ideas they genuinely find interesting, which is itself the ranking signal. Game-changing insights emerge as a byproduct.
Origin
Komoroske's own tactic, developed from open-source community-organizing techniques and first run at Google as a secret group he named 'Navel Gazers'. He has started eight or nine over roughly a decade. Related concepts he cites: Ken Stanley's novelty search, and Tim Urban's 'idea lab'.
Core principles
- 01Make the name slightly off-putting to outsiders so only intrinsically-motivated people join.
- 02Optional, secret, and consequence-free is what makes honest exploration possible.
- 03Yes-and is not soft: with limited time, people build only on what is genuinely surprising and valuable — that IS the rigour.
- 04Convergent building by many people on one idea is the signal that something real is there.
- 05It cannot be steered toward an outcome. Try to aim it and it stops producing.
- 06A community with one person talking is already dead — it just does not know it yet.
How to run it
- 1
Start with the smallest live seed
Find the four-ish people who already have generative conversations over lunch and formalize that into a group. Never launch into a big empty room — a cavernous space where nobody talks teaches people that this is a place where we do not talk.
Pro tip Keep the space too small and the time too short. Communities run on momentum, and scarcity creates it.
Watch out Do not try to convince someone who does not want it to want it. They will ruin the whole thing.
- 2
Name it so only the right people self-select
Give it a name that makes normal people wrinkle their nose — 'Navel Gazers', 'nerd club' — so that the only people who ask to join are the ones who want it for its own sake. That guarantees positive, yes-and energy from day one.
Pro tip The reaction you want on hearing the name is 'that sounds like a club for nerds' — answer: 'yes. Do you want in?'
- 3
Set the yes-and norms explicitly
State that this is a collaborative debate environment with yes-and norms. If someone posts an idea you think is actively dumb, you are free to simply not engage — nothing is being decided. If you do want to push back, do it by making it about you: 'so interesting, I'd never have thought to apply that lens; I typically apply this one — I wonder if it applies here.'
Pro tip Saying 'I wonder' makes the disagreement about you, not them, so the other person can choose whether to build on it.
Watch out One jerk can poop the party. Norms have to be set before scale, not after.
- 4
Dribble in new perspectives, one to three a week
Add members slowly, prioritising maximum difference of perspective — this is novelty search across viewpoints. Adding many at once lets a handful of strong personalities scramble the norms.
Pro tip Diversity of perspective is the raw material the group metabolizes into insight.
Watch out Five people with a particular personality arriving together can permanently break the culture.
- 5
Actively garden the conversation
Prune like a gardener: quietly tell Jeff his reply came out a bit strong to Sarah's idea and suggest opening with 'I wonder' next time. When someone DMs you a good thought, tell them to share it in the group, then respond enthusiastically in public — onlookers assume Sarah stuck her neck out unprompted and it worked, which makes risk-taking self-sustaining.
Pro tip It is not a secret — it just goes unobserved, and that is enough to establish the norm.
Watch out Ensure interesting conversation happens even when you, the facilitator, are absent. A group that only lives when you are present is not a community.
- 6
Run quorum-guaranteed live sessions and FOMO afterwards
Never propose something the group might not take up. Pre-canvass a few people ('I wonder if we should do a live conversation sometimes — would you come?'), get yeses, then announce it as an experiment. Guarantee quorum so it never looks dead. Afterwards, post notes thanking named people and calling out the best insight, so those who missed it feel FOMO and come next time.
Pro tip Never float an idea in the group that could land with silence — pre-load the yes.
Watch out A live session without quorum makes the whole community look dead and is very hard to recover from.
In the wild
Joining a team where twelve different groups worked on aspects of one problem domain and — in classic Google fashion — added up to less than the sum of their parts, with strategies that directly undermined each other. Komoroske judged that pairwise executive reviews on a problem this ambiguous would produce expensive pageantry that obscured more than it clarified. So he created a secret opt-in group instead.
→ A group people found intrinsically valuable that also spun off game-changing insights for the surrounding strategy. He has since started eight or nine such groups; several still run at Google.
Because members have limited time, they only build on ideas they find genuinely surprising and potentially valuable. So when many different people in the group start building on the same idea, that convergence acts as an organic ranking function on the idea space.
→ Rigorous prioritization emerges from a supposedly non-rigorous yes-and environment — and the winning ideas get told 'you should write that down'.
Common mistakes
Trying to steer the salon toward an outcome
The value comes from bottom-up emergence. The moment you aim it at a deliverable, participation becomes performance and the insights stop.
Launching big
A large room with no conversation trains people that this is a place where nobody talks. Momentum is everything: too-small space and too-short time is the correct starting configuration.
Recruiting the unwilling
One person who is there for the wrong reasons can kill the norms — 'it takes one person to poop a party'. Only add people who already roughly want it.
Is it for you?
Best for
Strategy leads, PMs and organizers facing a complex ambiguous cross-team problem where formal reviews would generate pageantry rather than clarity.
Not ideal for
Situations that need a decision by Friday. Salons produce insight on their own timeline and cannot be forced.
From the transcript
“so what I did was I created a secret group that I called Naval gazers with the original one”
“you say you set the Norms very explicitly and say this is a collaborative debate environment it says only yes and”
“productive way of doing that is saying oh it's so interesting I would never would have thought to apply that lens I typically would apply…”
“the third thing you do is you dribble in New Perspectives so every you know one to three a week”
“a community with one person talking is already dead you don't realize it yet”
“then what you do is you send fomo stuff afterwards”
From the episode
Thinking like a gardener not a builder, organizing teams like slime mold, the adjacent possible, and other unconventional product advice
Alex Komoroske (Stripe, Google)