Ask when to hire RevOps and you will get a number: four to five reps, or 25–50 employees, or somewhere between Series A and Series C. Those benchmarks are real, and they are also proxies for something more specific that is easier to spot.
What the benchmarks say
The published guidance clusters fairly tightly. Most Series A SaaS companies bring in their first RevOps manager somewhere around 25–50 employees, or once the sales team passes four or five reps. RevOps is most commonly introduced between Series A and Series C, depending on how complex the motion is.
Those work as a sanity check and poorly as a decision rule, because two companies at 30 people can have wildly different needs depending on whether they sell one product to one persona or four products through two channels.
Three signals that beat the benchmarks
Each of these tells you more than headcount does, and each is observable this week.
1. Two reps describe the same stage differently
Take five open deals. Ask two reps independently which stage each is in and why. If they disagree on more than one, your pipeline data has already stopped being comparable across people, which means the aggregate number has stopped measuring what you think it measures.
This is the single earliest reliable signal, and it usually appears well before anyone is talking about headcount. The fix is stage definitions with exit criteria, which is a week of work rather than a hire.
2. The weekly number requires an export
If producing the pipeline review involves downloading anything into a spreadsheet, you have an unstaffed reporting function. Someone is already doing RevOps, manually, at the end of the week, on top of their real job.
3. A fix has been "next sprint" for two months
Routing rules, a broken handoff, a field nobody updates. Everyone agrees it should be fixed. Nobody owns it. That is what an unstaffed function looks like from the inside: the work is visible, agreed, and permanently deprioritised against anything with a revenue number attached.
You do not decide to need RevOps. You notice that someone has been doing it badly, at nights, for a quarter.
Why sales leadership comes first
One sequencing point that most founders get wrong: in almost every case, hire the sales leader before the RevOps person.
RevOps builds the system that executes a strategy. With nobody setting that strategy (no agreed motion, no defined segments, no view on how deals should be run), your RevOps hire spends their first two quarters making those calls themselves, which is rarely the job they are best at and never the one you budgeted for.
The exception is a founder-led company where the founder is the sales leader and intends to stay that way for another year. RevOps first is defensible there: the strategy exists, it just lives in the founder's head, which is the problem RevOps solves.
The year before you can hire
Most companies feel this need twelve months before they can justify a $160k+ salary against another engineer or AE. Four options, honestly assessed:
| Option | Real cost | Where it breaks |
|---|---|---|
| Founder does it | Evenings and weekends | Stops the moment a quarter gets hard |
| Ops-minded rep does it | Part of your best rep's quota | You have quietly cut selling capacity to build systems |
| Sales leader absorbs it | Their strategic time | Selling pressure always wins over systems work |
| Fractional partner | A retainer, for a defined window | If they build something your team cannot run |
That last risk is the one to underwrite explicitly. Any external partner should be documenting themselves out of the job from week one, which is why our ninety days end with a written playbook your team keeps regardless of what happens next.
What to actually hire for
When you do hire, resist the temptation to recruit for analytics. The first RevOps hire's job is data integrity: whether the records are accurate, whether the definitions are shared, and whether the same event always produces the same result in the system.
Dashboards built on unreliable inputs are worse than no dashboards, because people act on them. Hire someone who finds that unglamorous work satisfying rather than someone who wants to build forecasting models on day one. The models can come once the inputs deserve them.
If you want an outside read on whether you are at the hire point or the fix point, the free GTM teardown will tell you, and you keep the findings either way.