Conviction Staffing
Staff speculative projects with people obsessed with the idea, not with whoever happens to be free.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 94%
Beykpour's biggest staffing pet peeve: large companies allocate people by availability, not by belief. On hard or speculative projects that is fatal, because every project requires willing the thing into existence, and only obsession generates that will. His rule is a cheat code — if the bet is risky, the entry criterion for the team is that they believe in it.
Origin
Learned the hard way by Kayvon Beykpour at Twitter, where a functional org with no single staffing decision-maker (and a CEO who wouldn't get into the weeds on staffing) repeatedly put unconvinced people on speculative features — most visibly hide replies.
Core principles
- 01Skill is table stakes; obsession with the idea is the differentiator.
- 02Belief makes people work harder, be more creative, and carry the ambition needed to will a project into existence.
- 03Building something from nothing is hard enough; building it with a team that doesn't believe in it is toxic.
- 04Belief must not become blindness — the team still has to be willing to see whether the thing is actually working.
- 05In functional orgs with distributed staffing decisions, this failure mode is the default, not the exception.
How to run it
- 1
Classify the project
Separate work into well-scaffolded core work (where availability-based staffing is fine) and speculative/risky work (where it is not). Only the second bucket needs this rule.
Pro tip Be honest: if the project has no proven demand and no existing surface, it is speculative, however senior the sponsor.
- 2
Screen for obsession before skill fit
Ask candidates for the team whether they actually want to learn whether this solves the customer problem. Staff people who are, in Beykpour's words, perhaps irrationally obsessed with the idea — while still willing to see truth.
Watch out A begrudging 'I'll go along with it' is a no. Passive dissent inside a speculative team drags everyone down.
- 3
Name a single staffing decision-maker
In a functional org, staffing decisions have no owner below the CEO — and the CEO won't debate them. Establish who decides, so belief-based staffing is enforceable rather than aspirational.
Pro tip This is one of the concrete arguments for a GM structure over a functional one.
- 4
Delay rather than backfill with the unconvinced
If the believers aren't available yet, wait for the right person instead of starting with whoever is free. Starting badly-staffed guarantees a false negative on the idea.
Watch out A speculative bet that fails with an unconvinced team teaches you nothing about the hypothesis.
In the wild
The hide-replies team did not even agree internally that the feature was a good idea or worth trying. An engineer told the PM that working on it would damage her career. The staffing had been driven by availability inside a functional org, not by belief in the hypothesis.
→ The feature shipped only through sheer force of will and enormous effort — and Beykpour drew the general rule from the pain.
Keith Coleman was extremely passionate about crowdsourced moderation at a time when most people thought it was a terrible idea. He was given the project because of that obsession, not because he had spare capacity.
→ Community Notes survived internal scepticism, launched, and outlived the entire leadership regime that started it.
Common mistakes
Staffing by availability
The default in large functional orgs is 'who is free?' rather than 'who is right?'. On speculative bets that is an unforced error, because motivation is the scarce input, not headcount.
Confusing obsession with denial
You need people irrationally committed to the idea AND still willing to see whether it is working. Belief without honest measurement is just a longer, more expensive failure.
Assuming a begrudging team will come around
A startup where the working team doesn't believe in the idea would never succeed. There is no reason to expect a different result inside a big company.
Is it for you?
Best for
Product and engineering leaders in matrixed or functional orgs who are launching speculative 0-to-1 initiatives inside a large company
Not ideal for
Well-understood maintenance, platform or optimisation work, where competent execution matters more than conviction
From the transcript
“you need to staff projects with the team of people that are well equipped from a skill set standpoint but more importantly have have an…”
“a very easy cheat code for an organization to employ is to say if you're going to work on something especially if it's speculative or…”
From the episode
Twitter’s former Head of Product opens up: being fired, meeting Elon, changing stagnant culture, building consumer product, more
Kayvon Beykpour