The Sidecar Model of Management
You don't take the wheel or hand it over — you ride attached, and you have to be invited in first.
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 92%
New managers default to two broken mental models: divide-and-conquer (you go there, I go here) or teaching-to-ride-a-bike (I hold on, then let go and hope you fly). Singhal argues the right model is the sidecar of a motorcycle — the other person drives, you're permanently attached, and your role is specific and advisory. He pairs this with a second rule: organizational power is not earned power, and like a vampire you cannot cross the threshold until you're invited.
Origin
Singhal's own framing, built from managing product orgs at Google, Credit Karma and Meta and from his broader claim that tech's defining failure is not training managers at all.
Core principles
- 01There is loose coupling between the skill of building things and the skill of teaching people to build — knowing how to make a car does not mean you know how to build the factory.
- 02The central question of management is how you share the steering wheel, not whether you hold it.
- 03New managers have organizational power, not earned power, and mistake the first for the second.
- 04Anything you say before you are invited in will just be annoying and will backfire.
- 05Once invited, pick one or two areas and genuinely partner — don't manage everything you can define as manageable.
How to run it
- 1
Reject the two default models
Consciously discard divide-and-conquer ('you take that, I take this, we meet up') and the bicycle model ('I hold the seat, then let go and hope'). Both resolve the steering-wheel question by avoiding it.
- 2
Take the sidecar seat
Adopt the posture: they drive the motorcycle, you're in the sidecar. You are attached whether either of you likes it or not, and your role is specific — giving counsel — not steering.
- 3
Ask to be invited in
Open with 'what can I help you with?' rather than defining what you will manage. You have to earn the right to be someone's manager no matter how senior you are.
Watch out A common answer is 'I don't need you, I wasn't excited about you as a manager, I don't need another layer between me and the CEO.' The correct response is to accept it — not to assert authority.
- 4
Wait in the sidecar
If you're refused, don't force it. Stay attached and available. One day they will need help; be there to say 'perhaps I can assist.'
- 5
On invitation, pick one or two areas and share the responsibility
When the invitation finally comes, don't sprawl. Choose a specific area or two, partner deeply on those, and make the shared ownership of them explicit.
Pro tip The three-part recipe: invitation, then specificity, then shared responsibility.
In the wild
Singhal starts managing someone who tells him plainly: I don't need you, I wasn't excited about you as a manager, I don't want another layer between me and the CEO. Rather than pulling rank, he accepts it — 'that's cool, because anything I say after that is just going to be annoying and it's going to backfire' — and waits in the sidecar.
→ The manager preserves the relationship and is positioned to be genuinely invited in the day the person actually needs help, at which point the partnership is real rather than imposed.
Singhal's diagnosis: in tech's first ~30 years, a good coder was assumed ready to manage, someone who could talk was assumed ready to product-manage, and anyone who shipped was assumed ready to tell people what to do. Almost no one was trained.
→ A thousand downstream problems — bias, missed development areas, poor coaching — and a generation of average managers who were promising ICs.
Common mistakes
Managing by threshold-crossing
Because you have organizational power, you start managing whatever you define as manageable. Uninvited counsel is annoying and backfires — you have to be let in.
Divide-and-conquer as a management model
It's coordination, not management, and it dodges the actual question of how the steering wheel is shared.
Assuming builder skill transfers to teaching
Being able to make a car does not mean you know how to build the factory that makes the car. The coupling is loose, and pretending otherwise is why the industry ships untrained managers.
Is it for you?
Best for
First-time managers, and experienced leaders inheriting a team or a skeptical senior report
Not ideal for
Crisis or turnaround situations where a direct, unilateral decision has to be made now and there is no time to be invited in
From the transcript
“I think it's more like the sidecar on the motorcycle where persons on the driving the motorcycle and I'm on the side car and whether…”
“what I find is that you're more like the vampire knocking on the door of someone you have to be invited in you just can't…”
“I generally think that it's this invitation picking specific and then making sure we're sharing the responsibility”
“it's the difference between if you can make a good car you must know how to make the factory that makes the car I don't…”
From the episode
Building a long and meaningful career
Nikhyl Singhal (Meta, Google)