Hire specialists where the work has a known shape and a known result. Bring in a generalist operator where the work is fuzzy, shifting, or sitting outside the org chart. Sorting the work first decouples seats from impact, and it lets you flex capacity up, down and sideways instead of running a new hiring process every time the business moves.
The Problem With Hiring Generalists "Because We Can't Afford Specialists"
It's a common thought that early teams hire generalists because payroll won't stretch far enough to put one person in one seat. Most founders are already living it. You are the generalist: you're sales, you're marketing, you're customer success, because there aren't enough hands and there isn't enough capital to buy more of them.
Then the first real hires come up and the logic flips. You stop thinking about hats and start thinking about problems. I have a specific problem, so I want a person who knows how to solve that specific thing. That instinct is right. It's also doing half the job, because budget is the wrong variable to sort on.
Three things go wrong when cost is the sorting rule:
- The generalist gets framed as a discounted specialist, so you expect less depth and you get less.
- You assume the problem will hold its shape long enough for a specialized role to stay correctly scoped.
- Every change in the business becomes a hiring event, because nobody on the team is built to absorb a change in the work.
What the Generalist Operator Skill Set Actually Covers
The word generalist carries a negative connotation. It sounds like someone running between functions, never going deep enough to affect change or to hold a thoughtful strategy. I've built a company entirely of generalist operators, internal and embedded, and that's not what the role is [1].
Depth in adaptability
The depth is real and it sits in adaptability. At the core, this kind of operator flexes across functions and operates without a fixed lane, which means they know how to build the infrastructure that lets a function run without them standing over it.
Judgment that doesn't come from a playbook
There's an innate ability to show up with discernment: to abandon what isn't working, to unlearn as fast as you learn, to stay curious while rolling up your sleeves and learning by doing. That's the part you can't write into a job description, and it's the part that holds up when the brief changes mid-quarter.
You've already met this person
If you know the chief of staff role, the proxy to the leader who handles administrative, strategic and hands-on work at once, you know the generalist operator. Most people assume chief of staff is the only way to get one onto a team. It's one way.
how We Do: Sort the Work Before You Sort the Hire
Before you open a role, sort the work. Five steps, and you can run them in an afternoon with a whiteboard.
- List the work, not the roles. Write out every stream of work the next two quarters require. Ignore titles completely at this stage.
- Mark each stream clear or fuzzy. Clear means you can describe what goes in, what comes out, and what good looks like. Fuzzy means the shape is still moving, or nobody has owned it before.
- Give the clear streams to specialists. Defined work rewards depth in a single craft, and you can measure it honestly from week one.
- Give the fuzzy streams to a generalist operator. This is the work hiding in the cracks between roles, the work that appears because the organization is growing or changing.
- Re-sort every quarter. Streams move between columns. Fuzzy work becomes clear once someone has built the infrastructure for it, and new fuzz arrives with every pivot.
"What's clear for the inputs and the outputs, you hire your specialists. What's not clear, you bring in your generalists." Britt Gage, founder of oAT
Specialist Roles vs Generalist Roles
| Dimension | Specialist hire | Generalist operator |
|---|---|---|
| Scope clarity | Defined before the hire, stable for a year or more | Defined loosely, expected to change |
| Best-fit work | A craft with a repeatable output | Work that crosses functions or has no owner |
| How success is measured | Against the function's own metrics | Against what the business needed that quarter |
| When priorities shift | The role is re-scoped, or a new role is opened | The person moves with the work |
| Resource cost of a change | A hiring cycle, onboarding, ramp | A conversation about where capacity goes |
| Common failure mode | Correctly scoped for a business that no longer exists | Spread across too many streams at once |
Neither column is the better column. The error is putting fuzzy work in the left column, which is how teams end up re-scoping a role twice in nine months and calling it a people problem.
The Org Chart Was Built for a Slower Century
The org chart was designed in the 1800s by a railroad engineer, and it was a genuine breakthrough: clear people in clear roles so that work had predictable results across a system too big to hold in one head [2]. That design assumed technology moved at a pace you could plan around.
| Design assumption | Industrial era | Your company now |
|---|---|---|
| Rate of change in tools | Measured in decades | Measured in weeks, sometimes days |
| Who defines the role | The chart, in advance | The work, as it arrives |
| What scale requires | Known people in known boxes | Capacity that can move |
| What's outside the chart | An exception | A standing condition |
That structure was built for a different century. The alternative is deciding, on purpose, which work belongs in a box and which work needs someone who can shapeshift as the organization changes.
"Suddenly you decouple the functional nature of people in seats in an org with the impact that the people are making." Britt Gage, founder of oAT
Sorting a Real Example: A Go-to-Market Push
Founders come to me all the time and say: I have a go-to-market need, I need to know how to scale, I need content, I need to stand up my sales ops. Can a generalist do that? Absolutely. Here's how the sort looks on a single GTM program.
| Work stream | Clear or fuzzy | Who takes it |
|---|---|---|
| Paid acquisition against a proven ICP | Clear | Specialist |
| Defining the ICP while the product is still moving | Fuzzy | Generalist operator |
| Standing up the CRM and the sales operating cadence | Fuzzy at first, clear once built | Generalist operator, then handoff |
| Outbound sequencing once messaging converts | Clear | Specialist |
| Building, training and supervising AI agents in the funnel | Fuzzy | Generalist operator |
| Relationship work with early customers | Fuzzy | Generalist operator |
A generalist does that work meaningfully. They build the operating infrastructure, build the agents, train the agents, babysit the agents, and they do the functional judgment work and relational work alongside it. Then, as the role changes, the person changes with it.
AI Keeps Pushing Work Into the Fuzzy Column
Work has always changed. What's different is the pace. New capabilities land by the month and sometimes by the hour, and nobody has settled what can be handed to an agent and what can't. That uncertainty shows up on your team as work with no obvious owner.
- Teams now include people and technology acting as people, which means new ways of working and new cultures to build.
- Somebody has to orchestrate the tooling and run the operations of it while the rest of the work still ships.
- The answer to "should an agent own this?" changes every quarter, so the owner of that question has to be comfortable being wrong on a short cycle.
- oAT covers this terrain as part of normal operator scope, alongside marketing, GTM, sales, business ops, customer success, product, community and chief of staff support [1].
What Changes When You Flex a Generalist Fractionally
When a generalist operator is on your team fractionally, which is how we work, capacity moves up, down and sideways into the places that need it. You spend less on hiring, because a shift in priorities doesn't require a new seat. You get the right team at the right time instead of the team you scoped last spring.
The practical effects:
- Fewer rehiring events. The person absorbs the change rather than the chart absorbing a new box.
- Infrastructure stays owned. Systems built during a push don't go dark when the push ends.
- Specialists stay specialized. Your best craft hires stop getting pulled into work that was never theirs, a pattern I've watched drive good people out of good roles [3].
An embedded operator joins your team and works in your tools, so the flexing happens inside your systems rather than across a vendor boundary.
A Short Checklist Before You Open the Role
Run these four questions before you write the job description.
- Can I describe the output of this role in one sentence that will still be accurate in twelve months? If yes, specialist.
- Has this work had three different owners in the last year? If yes, it's fuzzy, whatever the title said.
- Am I hiring for a problem, or for a season of problems? Seasons go to generalists.
- If this work doubled next quarter, what breaks first: the person, or the system they never had time to build?
"I am not suggesting a world where every person on your team is a generalist, but we lived in a world where every person on the team was specialists, and there wasn't enough slack in the system, which is why the generalist operator existed." Britt Gage, founder of oAT
You can design for that slack now instead of discovering you needed it.
FAQ
Should an early-stage startup ever hire a specialist?
Yes. A specialist is the right hire whenever a stream of work has a defined brief, a measurable output, and enough stability to stay that way for several quarters. Paid acquisition against a proven customer profile, a specific engineering discipline, or reg