Stop Posting Jobs, Start Posting Problems

Stop Posting Jobs. Start Posting Problems.

Pinned notes with one labeled problem in focus I’ve been circling a recruiting model that feels almost embarrassingly obvious once you say it out loud:

Companies shouldn’t post jobs. They should post problems they want solved.

Not “Senior Backend Engineer (Go, Kubernetes, 7+ years).” Not “Product Manager, Growth.” Not another wishlist of keywords plus a salary band that’s either fake or “DOE.”

Just the problem. The real thing that’s actually hurting.

And then candidates don’t apply with resumes. They respond with proposals—short, structured write-ups that double as a thinking sample and an interview. If there’s a fit, the company creates the role around the person.

That’s the whole model. It’s also a direct indictment of how broken hiring has become.

What “Problem Posts” Look Like

A job description is an abstraction. A problem statement is reality.

A good problem post sounds like:

“Churn is high among users who onboard via mobile. We think it’s an onboarding issue but we’re not sure where the drop-off is.”
“We ship too slowly. Release process is brittle and everyone is afraid to touch it.”
“We want to validate a new AI feature in 30 days without derailing the core roadmap.”
“Our customer feedback is unstructured and we can’t turn it into decisions. We need a system.”

Even better if it includes constraints and context:

timeline
what’s been tried
what data exists
what “success” looks like
what can’t change (budget, team size, tech stack, compliance)

This is what teams actually talk about internally. This is what founders complain about at dinner. This is what makes or breaks a quarter.

So why do we translate it into “We are seeking a rockstar…”?

Proposals as Interviews (Without the Theater)

Resumes are mostly metadata: past titles, brand names, keywords. Interviews are often performance: confidence, storytelling, social calibration, and luck.

A proposal, done right, is higher signal because it forces contact with reality. It shows:

how someone frames ambiguity
what they prioritize first
what questions they ask before acting
how they think about constraints
what risks they see that others miss
whether they have taste

It also shifts the candidate’s posture from “please pick me” to “here’s how I’d attack this.” That’s closer to how high-leverage work actually operates.

The proposal doesn’t need to be long. In fact, it shouldn’t be. Think:

a few paragraphs of understanding + approach
a short plan (phases, milestones)
key questions / missing info
risks and tradeoffs
what they’d need (access, tools, decision-makers)

It’s basically an interview that starts with substance instead of vibes.

Why Job Descriptions Are Fake Precision

The biggest lie in hiring is that the company knows what it needs.

Job posts pretend the organization has already figured out:

the exact responsibilities
the required skills
the seniority level
the ideal background
the salary band
the scope boundaries
the internal decision-making model

But most teams—especially growing teams—are guessing. They’re trying to solve a real problem, and the “role” is their best attempt to describe a solution shape.

The problem is: guesses calcify. Once you name the role, add requirements, and set filters, you start rejecting people who could solve the problem simply because they don’t match the invented shape.

Problems don’t have that issue. They’re what they are. A key fitting a uniquely shaped lock

Why Interviews Are Terrible Proxies for Modern Work

A lot of hiring is still optimized for a world where:

tasks are predictable
roles are stable
output is easy to measure
organizations are slow-moving

But modern high-leverage work is messy:

the real constraints are hidden
goals shift mid-stream
tradeoffs matter more than “best practices”
communication is part of execution
judgment is everything

Standard interviews barely touch that. They reward:

social performance under pressure
rehearsed narratives
charisma and confidence
pattern-matching to what interviewers expect

Meanwhile, the skills you actually want—clarity, prioritization, decision-making under uncertainty—show up much more clearly when someone is forced to grapple with a real problem statement.

If you care about outcomes, you should be hiring for how people think, not what they’re called.

Great People Don’t Want “Jobs” (They Want Leverage)

The best operators, builders, and thinkers often don’t fit cleanly into a title box:

they’re cross-functional
they’ve done weird combinations of things
they’re allergic to bureaucracy
they want autonomy and impact
they don’t want to cosplay in a role that doesn’t match reality

A job post asks them to pretend: “Apply as a ‘Growth PM’ and then we’ll see if you can really fix retention.”

A problem post invites them to engage honestly: “Here’s the retention problem. Here’s how I’d approach it.”

That’s a much more respectful transaction. It treats capability as something you demonstrate, not something you claim.

This Isn’t a Job Board. It’s a Role-Creation Engine.

If you build this as “a job board but with problems,” it’ll collapse into the same old patterns. The differentiation is what happens after the proposal.

The point is:

Company posts a problem.
Candidates submit proposals (structured, time-bounded, non-exploitative).
Company talks to the most compelling thinkers.
They shape the role together:
- scope
- success metrics
- autonomy level
- contract vs full-time vs fractional
- what support exists internally
- what “winning” looks like in 30/60/90 days

Most hiring flows stop at “we chose a person for a predefined box.” This flow admits the box was always imaginary and builds the right one after you find the person.

That’s the product.

The Line You Cannot Cross: Proposal ≠ Free Work

If proposals become unpaid consulting deliverables, the model dies instantly.

You need hard norms that protect candidates:

proposals are approach, not execution
no “deliver a full roadmap” requests
no “audit our entire codebase” as an application step
strict length limits (or time-box expectations)
problem posts may need a fee or moderation to deter spam

A good proposal should feel like: “Here’s what I’d do, here’s what I’d watch for, here’s what I’d ask, and here’s how I’d de-risk this quickly.”

It should not feel like: “I just did two days of work to earn a maybe.”

Problem Quality Is Everything

Bad problem posts will ruin the marketplace.

Bad looks like:

“We need growth.”
“We want to use AI.”
“Fix our tech debt.”
“Make the product better.”

Good looks like:

outcome-focused
time-bounded
context-rich
honest about constraints
clear about what’s known vs unknown

If companies can’t articulate problems, that’s already a signal—but it’s also a solvable product issue. The platform could guide them with templates:

What’s the current state?
What’s the pain?
What have you tried?
What constraints exist?
What does success look like in 4–8 weeks?

This is less “write a job description” and more “write a decision memo.” Which is the point.

Who This Works For (First)

This model fits best where problems are concrete, stakes are real, and leverage matters:

Great early targets:

startups
small teams
founder-led orgs
product, engineering, data, design, strategy roles
fractional / contract-to-hire pathways

Harder early targets:

high-volume hiring
heavily regulated roles with fixed credentialing
very junior roles (where proposals could become noise)

Starting narrow isn’t a weakness; it’s focus. High-signal marketplaces win by being opinionated.

The Hard Question: How Is This Not Just “Upwork With Better Branding”?

If the platform becomes “post task, get bids,” it’s over. That’s not recruiting; that’s commoditized gig work.

The defense is structural:

curated, well-scoped problems (not random tasks)
proposals that are constrained, comparable, and high-signal
emphasis on role creation, not one-off completion
success measured by durable matches, not transaction volume

If you get those mechanics right, it’s not a freelancing site. It’s a hiring system that starts with reality instead of theater.

Conclusion

Job postings are a clumsy translation layer between what companies actually need and what candidates can actually do. Posting problems and evaluating proposals skips the performative parts and surfaces judgment, clarity, and execution instinct. The win isn’t just better screening—it’s building roles around real work and real people. If hiring is supposed to be about outcomes, it should start with the outcome you want.