RevOps vs sales ops vs marketing ops: which do you need?

Sales ops serves the sales team. Marketing ops serves demand generation. A CRM manager serves the system of record. A RevOps manager owns the numbers across all three and arbitrates between them. A GTM engineer builds the automation underneath. Choose by symptom: unreliable reporting points to RevOps, blocked throughput points to GTM engineering, and a broken CRM points to neither.

The short answers

• What is sales operations? The function that makes a sales team’s process repeatable: territories, quotas, forecast hygiene, deal desk, compensation administration and the reporting the sales leader plans against. Its boundary is the sales org.
• Is RevOps the same as sales ops? No. RevOps is the same discipline with the boundary moved to the whole revenue funnel, which means it also arbitrates definitions between marketing, sales and customer success. Below about 150 employees the two titles usually describe one job.
• Which ops role should you hire first? Almost always the generalist RevOps manager, because a specialist inherits definitions that do not yet exist.
• Who does RevOps report to? Most commonly the CRO or VP Sales at Series A and B, and increasingly the CEO, COO or CFO from Series C. No published survey establishes clean percentages, and anyone quoting them is quoting an opinion.
• What does a GTM engineer add? Execution capacity once the process is already defined. They ship systems that run without them.
• What is the ratio of ops headcount to sellers? Published figures range from 1 ops person per 10 revenue staff to 1 per 35 quota carriers. Both are practitioner heuristics. Neither comes from a sampled study.

What is sales operations

Sales operations is the internal function that makes a sales team’s process repeatable and its numbers trustworthy. It owns territory design and account assignment, quota setting and capacity modelling, forecast cadence and hygiene, pipeline inspection, deal desk and approvals, compensation plan administration, and the sales-facing configuration of the CRM.

The job exists because selling time is scarce and administrative load is not. In the seventh edition of the Salesforce State of Sales report, a survey of 4,050 sales professionals across 22 countries conducted in August and September 2025, respondents reported spending 40% of their time selling and 60% on everything else, with 17% going to data entry alone. A quarter of that sample worked in sales ops. The same research puts the average sales team on eight tools and finds 42% of reps saying the tool count overwhelms them.

Sales ops is the function that attacks those numbers. It is measured on whether the sales organisation can plan: whether the forecast holds, whether territories are fair enough that reps stop arguing about them, whether commission statements arrive without disputes, whether a deal can move from verbal agreement to signature without four people improvising.

Gartner’s guidance to sales operations leaders groups the remit into mission definition, financial planning, workforce planning, process optimisation, technology strategy and risk management. That is a wide brief, and the absences matter: nothing about marketing’s lead definitions, nothing about renewal or expansion motions, nothing about the customer success handover. Those omissions are the whole reason the RevOps title exists.

Sales operations is also the oldest of these five titles by a decade or more, which matters when you read a CV. A candidate with fifteen years in sales ops at large enterprises has usually specialised deeply into one or two of those areas. A candidate with three years in sales ops at a Series B company has usually done all of them badly at once, which is a different and sometimes more useful profile.

Is RevOps the same as sales ops

No, and the difference is structural rather than cosmetic.

Sales operations answers to the sales leader and optimises the sales team’s funnel. Revenue operations answers to somebody above any single revenue function and optimises the whole path from first touch to renewal. That produces three concrete differences.

Scope of funnel. Sales ops owns the stages a seller touches. RevOps owns the stages before the seller (lead capture, scoring, routing, MQL to SQL conversion) and the stages after the close (onboarding handover, expansion, churn signal, renewal forecast). The most expensive leaks in a B2B funnel sit at the joins, and only a function that owns both sides of a join can fix one.

Whose definitions win. When marketing counts an MQL one way and sales counts a qualified opportunity another way, sales ops cannot settle it, because sales ops reports to one of the two parties. RevOps can, because it does not. Definitional arbitration is the single duty that most cleanly separates the two roles, and it is the one most job descriptions leave out.

Who the internal customer is. Sales ops has one internal customer and can therefore be judged on that customer’s satisfaction. RevOps has three or four internal customers with conflicting interests and must be judged on business outcomes instead. That is a harder job to specify and a much harder one to interview for, which is part of why RevOps job descriptions attract the wrong applicants so reliably.

The shift is real and it is measurable. Forrester’s survey of B2B revenue operations, based on 316 respondents, found centralised revenue operations rising from 15% to 40% of organisations across two years, with centres of excellence rising from 18% to 37% over the same period. Gartner predicted in May 2021 that 75% of the highest growth companies in the world would deploy a RevOps model by 2025, though that press release discloses no sample size or methodology, so treat it as a directional call. The title also moved fast in the labour market: LinkedIn’s fastest-growing US roles for 2023, reported by Fortune, placed Head of Revenue Operations first, drawn from job titles posted between 2018 and 2022.

Here is the uncomfortable part. Below roughly 150 employees, the two titles usually describe the same job. One person owns every system, every definition and every report, and whether their business card says sales ops or revenue operations is decided by the hiring manager’s own background and by which word pulls better applications. Pretending otherwise in a job advert produces a mismatch on day one, because the candidate reads “revenue operations” as a cross-functional mandate and then discovers they report to the VP Sales and are expected to rebuild the commission spreadsheet.

What marketing operations owns

Marketing operations owns the demand side: the marketing automation platform, the lead lifecycle from anonymous visit to routed record, scoring models, campaign taxonomy and attribution, form and landing page infrastructure, list hygiene, consent capture, and the data protection obligations that come with holding marketing contact data.

The function grew because the tooling did. Scott Brinker and Frans Riemersma counted 15,384 marketing technology products in the State of Martech 2025, up 9% on the 14,106 they mapped a year earlier. Somebody has to decide which of those a company buys, how it connects to the CRM, and what happens to a record that touches three of them in a week. That somebody is a marketing ops manager, and the role has professionalised enough to sustain its own research: the State of the Marketing Ops Professional study surveyed more than 600 practitioners across 20-plus industries in 2024.

Attribution and compliance separate marketing ops from RevOps in practice. Attribution is a marketing ops speciality and a RevOps headache: building a model that survives contact with a finance team is a distinct skill from arbitrating between functions. On compliance, consent records, preference centres, suppression lists and lawful basis for processing sit with marketing ops in most companies, and a RevOps generalist who has never run a marketing database will not think about them until something goes wrong.

The overlap is genuine at lead routing. Both roles claim it, both roles can build it, and when both exist without an explicit decision, routing rules end up maintained in two places and diverge within a quarter.

What a CRM manager does

A CRM manager administers and develops the system of record. Object model and field architecture, page layouts, permissions and sharing, validation rules, sandbox and release management, integration monitoring, data quality remediation, and day-to-day user support and enablement.

The distinction that matters when you are choosing between this title and RevOps is what each is judged on. A CRM manager is judged on whether the system works: uptime, release quality, data integrity, ticket resolution, whether an admin change broke a report. A RevOps manager is judged on whether decisions got better. Those are different jobs and they attract different people, though the same person can hold both at small scale.

At scale they visibly separate. A team analysis of Yelp’s Salesforce organisation describes 3,000 Salesforce users supported by a 30-person core team containing five Salesforce administrators, two architects, thirteen business systems analysts, three CRM product owners and three engineers, with a further eighteen revenue operations specialists alongside it. Five admins to 3,000 users is one administrator per 600 seats, against the more commonly quoted enterprise practice of seven to ten admins per 1,000 users. The spread between those two figures is a fair warning about what published staffing ratios are worth, which is a subject the ratio section below returns to.

The hiring failure specific to this role is title drift in the other direction. Companies advertise for a RevOps manager, describe a CRM administration job, pay a RevOps band, and then wonder why the hire never produces a forecast anyone trusts. Read the responsibilities, not the title, and if the first six bullets are configuration tasks, you are hiring a CRM manager.

Where the GTM engineer fits

A GTM engineer builds and owns automated revenue systems across data foundation, data modelling and data activation, which is covered in full detail in the pillar on what a GTM engineer actually does.

For the purpose of choosing between roles, the distinguishing properties are narrow. They work in code and APIs, not just in configuration screens. They are judged on whether a system still runs six months later without them touching it. And they consume definitions instead of producing them, which is the property that fixes their position in the hiring order.

The role is young and growing quickly. GTME Pulse’s analysis of the GTM engineer job market examined 3,342 postings collected between January 2024 and February 2026 alongside a survey of 228 practitioners in 32 countries, and reported 205% year-on-year growth in postings between 2024 and 2025, a median advertised salary of $132,000, and Clay named as a required tool in 69% of adverts. That same analysis notes that reporting lines for the role vary widely, with GTM engineers sitting under sales, marketing, RevOps or engineering depending on the company. That variance is not a sign of an immature market so much as a sign that the role attaches to whichever function has the throughput problem.

The five role comparison

This is the table to read before writing a job description. Every row is a question a hiring manager should be able to answer about the role they are opening, and a role that fails three or more rows is not yet a role.

	Sales ops	Marketing ops	CRM manager	RevOps manager	GTM engineer
Owns	The sales team’s process, territories, quotas, forecast cadence	The demand funnel, MAP, scoring, attribution, consent	The system of record: objects, permissions, releases, data quality	Definitions, reporting and process across the whole revenue funnel	The automated systems that execute the process
Ships	A forecast and a comp plan the sales leader can run on	A lead lifecycle that routes correctly and reports honestly	A stable, well-governed CRM other teams can build on	One agreed set of numbers and the process that produces them	Infrastructure that runs without them
Judged on	Whether the sales org can plan and get paid correctly	Whether marketing spend can be traced to revenue	Whether the system works and stays working	Whether decisions got better	Whether it still works in six months
Fails by	Becoming the sales leader’s report-writing service	Building attribution nobody outside marketing believes	Optimising the system and ignoring the business question	Saying yes to everything and arbitrating nothing	Building something clever and brittle
Typical background	Sales ops, finance, analytics, consulting, ex-AE	Demand gen, martech agency, campaign management	Salesforce or HubSpot admin, business systems, IT	Sales ops, analytics, consulting, strategy, finance	Ops, growth, data engineering, software, technical founder
Reports to	VP Sales or CRO	VP or CMO Marketing	RevOps, IT, or business systems	CRO at Series A and B, CEO, COO or CFO later	RevOps, sales, marketing or engineering, depending on the bottleneck
Typical salary band	$85k to $145k	$85k to $150k	$80k to $135k	$100k to $175k	$110k to $200k
When you need it	Sellers are arguing about territory and comp, forecast slips monthly	Lead volume outgrew manual routing, spend needs justifying	Two admins have made conflicting changes and reports broke	Nobody agrees what a qualified lead or a closed deal is	Process is agreed but manual execution caps volume
What breaks if you skip it	Reps sell around the process, comp disputes eat leadership time	Leads sit unrouted, marketing budget cannot be defended	Data quality degrades silently until a board report is wrong	Every function reports a different number and planning stops	Growth is capped at what humans can do by hand

Salary bands are indicative individual contributor ranges taken from the detail on what each of these roles pays, which sets out how bands move with seniority, sector and location. Treat the column as a relative ordering. It is not an offer.

Where the roles genuinely overlap

Most role confusion is not confusion. It is a real overlap that nobody has assigned. These are the responsibilities that sit legitimately in two roles at once, and each one needs an explicit owner the moment both roles exist.

Responsibility	Claimed by	Also claimed by	Give it to	What happens when nobody owns it
Lead routing rules	Marketing ops	RevOps	Marketing ops builds, RevOps signs off the definition	Rules maintained in two systems, leads assigned twice or not at all
CRM object and field changes	CRM manager	RevOps	CRM manager, under a change process RevOps sets	Field sprawl, reports that silently exclude records
Forecast reporting	Sales ops	RevOps	Sales ops produces, RevOps owns the methodology	Two forecasts in the same board pack
Attribution model	Marketing ops	RevOps	Marketing ops builds, RevOps agrees the revenue definition	Marketing reports pipeline finance does not recognise
Data enrichment and list building	GTM engineer	Marketing ops	GTM engineer for the pipeline, marketing ops for the consent layer	Duplicate contacts, enrichment credits burned twice, consent gaps
Outbound sequencing infrastructure	GTM engineer	Sales ops	GTM engineer builds, sales ops owns rep adoption	Sequences nobody uses, or reps building shadow sequences
Warehouse models and reverse ETL	GTM engineer	Data or analytics	GTM engineer for activation, data team for the source models	Two versions of the same metric with different logic
Compensation plan administration	Sales ops	Finance	Sales ops calculates, finance approves and pays	Commission disputes, and a quarter spent reconciling them

The right-hand column repeats itself. Almost every failure is a duplication failure, not a capability failure. Two competent people building the same thing in two places produces a worse result than one competent person building it once, and it costs twice as much.

Which role fixes which symptom

Start from what is broken. The title follows.

Symptom	The role that fixes it	The hire people make instead	What to test before you hire
The forecast is wrong every month and nobody can say why	RevOps manager	A sales analyst, who reports the same wrong number faster	Give them last quarter’s pipeline export and ask what they would refuse to include
Marketing and sales report different pipeline numbers	RevOps manager	A BI tool, which produces a third number	Ask them to write the MQL and SQL definitions they would propose
Leads take two days to reach a rep	Marketing ops	A GTM engineer, who automates a routing rule nobody agreed	Ask them to map the current routing path end to end
Reps spend an hour a day on CRM admin	Sales ops or CRM manager	A RevOps manager, who is overqualified and bored by month four	Ask which three fields they would delete and what would break
Two admins made conflicting changes and reports broke	CRM manager	A contractor, who fixes it and leaves with the context	Ask how they run a release and what their rollback plan is
Territory arguments consume every leadership meeting	Sales ops	A new sales leader, who inherits the same map	Ask them to critique your current territory design in writing
You know your target accounts but cannot reach them at volume	GTM engineer	An SDR, which buys throughput without buying repeatability	Ask for a system they built that someone else now operates
Enrichment and list building eat a person’s whole week	GTM engineer	A data vendor, which relocates the work instead of removing it	Ask them to describe the last pipeline they built and its failure modes
Attribution cannot survive a finance review	Marketing ops	An attribution vendor, which sells a model finance also rejects	Ask what they would tell the CFO the model cannot prove
Commission statements are disputed every month	Sales ops	Compensation software, which encodes the same ambiguity	Ask them to find the ambiguity in your current plan document

Look at the third column. Hiring a GTM engineer to fix a routing delay is the most common expensive mistake in this list, because the delay is usually definitional and automation makes an undefined process fail faster. Hiring a sales leader to fix territory arguments is the most common political one, because it treats a design problem as a management problem.

Who does RevOps report to

There is no reliable published percentage for this, and the sources that appear to offer one are giving advice in place of data. RevOps Co-op’s guidance on the question recommends the CRO for a VP of Revenue Operations, with COO or CFO acceptable at younger companies, and offers no survey behind it. Forrester’s 316-respondent study measured whether the function was centralised, which is a different question from who signs the org chart. So what follows is a practical account of what each line optimises for, drawn from live searches instead of a survey.

Reporting to the CRO or VP Sales. The default at Series A and B. It gets the function proximity to the number and fast decisions, and it is the easiest line to fund because the budget already exists. The cost is arbitration. A RevOps manager who reports to sales cannot credibly settle a dispute between sales and marketing, and marketing will read every ruling as partisan. If you choose this line, expect to keep marketing ops separate under the CMO, and expect definitional disputes to escalate to the CEO.

Reporting to the CEO or COO. The line that makes the arbitration mandate real. It works when the CEO will actually adjudicate and fails when RevOps becomes an unowned function with no budget and no advocate. It usually appears somewhere between 150 and 400 employees, when the cost of three functions reporting different numbers exceeds the cost of another direct report.

Reporting to the CFO. Increasingly common in PE-backed and capital-efficient companies. It buys rigour, a clean line into planning, and a forecast finance already believes. It costs enablement: a RevOps function under finance drifts toward reporting and control, and away from process improvement that sellers experience as helpful. Candidates notice this in interviews and some will decline for it.

Reporting to a Chief Revenue Officer who owns marketing too. The cleanest arrangement of the four when it exists, because the arbitration happens inside one leader’s remit. It is rare below Series C.

Stage	RevOps reports to	Sales ops reports to	Marketing ops reports to	CRM manager reports to	GTM engineer reports to
Seed	Founder or CEO	Does not exist yet	Does not exist yet	Does not exist yet	Rarely exists
Series A	CRO or VP Sales	Same person as RevOps	Head of Marketing, often part time	Outsourced or a partner agency	Founder, CEO or head of growth
Series B	CRO, occasionally COO	Head of RevOps	CMO, dotted line to RevOps	Head of RevOps	Head of RevOps or head of growth
Series C and beyond	CEO, COO or CFO	VP RevOps	VP RevOps or CMO	VP RevOps or business systems	VP RevOps, sometimes engineering

The dotted line in the Series B row is where most structural arguments happen. Marketing ops usually keeps a solid line to the CMO because campaign execution is a marketing capability, while conceding a dotted line to RevOps for data definitions and lifecycle stages. That arrangement works. Moving marketing ops fully under RevOps at Series B tends to slow campaign execution enough that marketing builds a shadow ops capability within two quarters.

Revenue operations team structure by company stage

Ops appears later than founders expect and grows more slowly than sales. The table below describes what we see across live searches; it is not a surveyed benchmark. The ranges are wide because sales model matters more than headcount: a product-led company with self-serve revenue needs data and lifecycle work far earlier than an enterprise sales company of the same size.

Stage	Company headcount	Typical ARR	Ops headcount	Titles present	What is still outsourced
Seed	10 to 30	Under $1m	0 to 0.5	Nobody, or a founder and a part time contractor	CRM setup, reporting build, anything requiring a specialist
Series A	30 to 80	$1m to $5m	1	RevOps manager, generalist, hands on keyboard	Salesforce or HubSpot development, attribution, comp modelling
Series B	80 to 250	$5m to $20m	2 to 4	Head or senior manager of RevOps, plus a CRM or systems specialist, plus marketing ops under the CMO	Data warehouse build, complex integrations
Series C and beyond	250 to 800	$20m to $75m	5 to 10	VP RevOps, sales ops, marketing ops, CRM or systems, analytics, first GTM engineer	Little, though specialist Salesforce development often stays with a partner

The two configurations most companies actually run are these.

The two-person ops team, typical at Series B. One generalist RevOps manager and one systems specialist. The RevOps manager owns definitions, the forecast, the reporting layer and the relationship with sales and marketing leadership. The specialist owns the CRM, integrations, data quality and the ticket queue. The split works because it separates thinking work from build work, and it fails when the RevOps manager is pulled into the ticket queue during a busy quarter and never gets out. If you run this shape, protect the generalist’s calendar explicitly or you have bought two systems administrators.

The six-person ops team, typical at Series C. A VP or Director of RevOps, a sales ops manager, a marketing ops manager (often still solid-lined to the CMO), a CRM or business systems manager, an analyst, and a GTM engineer. The VP owns definitions and the planning cycle. The three functional managers own their funnels. The analyst owns reporting so the managers do not spend their weeks in a BI tool. The GTM engineer owns the automation layer across all of it and takes work from every other seat. This is the first configuration where the GTM engineer role clearly pays for itself, because there are now four other people producing defined processes for them to automate.

The same failure patterns show up in both shapes. Hiring the analyst last means the senior people spend their first year building reports. Hiring the GTM engineer first means automating processes that change three months later. Hiring a second generalist instead of a specialist produces two people with overlapping remits who quietly divide the company between them.

At what headcount does each role first appear? A single ops generalist typically arrives between 25 and 50 employees, or earlier in product-led businesses. A dedicated marketing ops hire arrives when marketing exceeds four or five people, which is usually 60 to 100 employees. A dedicated CRM or business systems manager arrives when the CRM has more than one integration and more than one person permitted to change it, commonly 100 to 200 employees. Sales ops separates from RevOps when there are enough sellers that quota and territory work is a full-time job, usually 15 to 25 quota carriers. A GTM engineer arrives when the process is stable and the constraint is execution volume, which can be at any headcount above about 50 and is driven by go-to-market motion rather than size. Whether you should be hiring any of them yet is a separate question, covered in what to look for in a first ops hire.

Ops headcount ratios, and where they come from

Anyone planning an ops team eventually searches for a ratio. The ratios exist. They do not agree with each other, and none of them come from a sampled study.

Claimed ratio	Source	Evidence base
1 RevOps person per 10 revenue staff across sales, marketing, AM and support	Kim Castlemain and Amanda Chandler, quoted by QuotaPath	Practitioner opinion. No study cited
1 per 15 or 20 at larger revenue organisations	Jamie Carney, quoted in the same piece	Practitioner opinion. No study cited
1 FTE per 25 to 35 quota carriers, or 1 per 30 to 50 total GTM employees	The Pedowitz Group staffing guide	Explicitly labelled “TPG Planning Heuristics”. No study cited
Grow the team when there is more than 40 hours of revenue-driving work	Eddie Reynolds, quoted by QuotaPath	Practitioner opinion, stated as a workload test rather than a ratio

The spread between the first and third rows is a factor of three. A 120-person GTM organisation would need twelve ops staff on the first figure and four on the third. Both numbers are published by credible people in the field. Neither is research, and both are stated as such by their authors if you read the source rather than the citation.

The published SaaS benchmark data does not resolve it either. Benchmarkit’s 2025 B2B SaaS Performance Metrics report, based on 583 participating companies, reports ARR per full-time employee of $240,000 in the $50m to $100m band and $283,379 above $100m, and a median general and administrative expense of 24% of revenue. Useful figures, and none of them break operations out as a line. The US Bureau of Labor Statistics has no occupational code for revenue operations at all, so there is no government wage or employment series to fall back on.

Treat the same caution as applying to the impact claims. The most widely repeated statistic in this field comes from BCG’s 2020 paper on go-to-market operations in B2B by Federico Fabbri and colleagues, which cites 100% to 200% increases in digital marketing return on investment, 10% to 20% increases in sales productivity and 30% reductions in go-to-market expenses. The paper names no companies, no sample size, no measurement period and no methodology. It is a consulting point of view, and it is quoted across the industry as though it were a finding.

The workload test in the last row of the table is the one to use. Count the hours of revenue-affecting ops work currently going undone or being done badly by someone whose job it is not. When that exceeds a full week, you have a hire. That is unsatisfying next to a clean ratio, and it is the only one of the four with a defensible basis. It also connects to the harder budget question of whether the next headcount should be ops at all, which is worked through in RevOps hire or another AE.

Why legibility has to come before scale

A RevOps manager makes the business legible. A GTM engineer makes it scalable. Legibility has to come first, because automating an undefined process produces fast, well-instrumented chaos.

The mechanism is worth spelling out, because the abstract version sounds like a slogan and the concrete version changes hiring decisions.

Take a company with 40 sellers and no agreed definition of a qualified opportunity. Marketing counts a demo request. Sales counts a demo attended with budget confirmed. Both numbers are in the CRM, and both are used in the same board pack. Now hire a GTM engineer. They build an enrichment pipeline, a scoring model and an automated routing system that assigns qualified opportunities to the right seller within ninety seconds. It works. It is genuinely good engineering. And it now applies one of the two competing definitions, at machine speed, to every record in the business, which means the disagreement between marketing and sales is no longer a monthly argument in a meeting. It is encoded, invisible, and running four thousand times a week.

Unpicking that is harder than building it was. The engineer has to be redirected onto rework, the seller trust that was already thin gets thinner, and the person who could have settled the definition in week two is hired in month nine into a mess.

Reverse the order and the same money buys something else. The RevOps manager spends six weeks agreeing definitions, publishes them, rebuilds the reporting layer on top of them, and produces a forecast the CEO stops arguing with. The GTM engineer arrives to a specification. Their first system encodes a definition the business has already agreed, so it survives contact with the next quarter.

There is one honest exception. If your process genuinely is agreed, documented, and stable, and the constraint really is manual execution volume, then GTM engineering first is correct and RevOps first is a delay. That case is rarer than founders think, and it is testable: ask three leaders to write down the definition of a qualified opportunity separately, then compare. If the three answers match, you have the exception. In most companies they do not match.

How to tell which role a candidate actually is

Titles do not survive contact with reality. Someone titled RevOps manager at a 60-person company has often been a CRM administrator. Someone titled sales ops analyst at a 900-person company has sometimes been running a genuine revenue operations function under a legacy title. The interview has to establish which.

The single best question stays the same: what did you ship last quarter, and who operates it now?

A RevOps candidate names a process, a definition or a reporting layer, and says other people use it. A GTM engineer names a system and says it runs without them. A CRM manager names a release or a data remediation project and can tell you the rollback plan. A marketing ops candidate names a lifecycle or an attribution change and can tell you which stakeholder objected. A candidate who cannot name anything they shipped has been maintaining, and maintaining is a real job that is not the one you are hiring for.

Four more probes separate the profiles reliably.

• Ask what they would refuse to automate, and why. GTM engineers with judgement have a clear answer involving process instability or a compliance boundary. Candidates without judgement say nothing is off limits.
• “Walk me through a definition you changed and who fought you.” RevOps candidates have a story with named opposition and a resolution. Sales ops candidates often have a story where the sales leader simply decided. Candidates with neither have never arbitrated anything.
• How do they handle a request from a sales leader that would break reporting for everyone else? The answer separates a service function from an owning function.
• Ask for the last thing they built that failed. CRM managers and GTM engineers both have technical failure stories and they sound different. One is about a change that broke something downstream. The other is about a system that worked until an API changed or a volume threshold moved.

None of this is reliable through a CV screen, and it is only partly reliable through an unstructured interview, which is why the assessment work sits where it does in the wider hiring process for a RevOps manager.

The order of the hires matters more than the titles

Every one of these five roles is worth hiring in the right circumstances, and the wrong one costs a year. The sequence that works in most companies is: a RevOps generalist who settles definitions and builds the reporting layer, then a systems specialist who governs the CRM, then marketing ops when demand volume outgrows manual routing, then a GTM engineer once there is a stable process worth automating, then the split of RevOps into sales ops and marketing ops as each funnel gets big enough to need a full-time owner.

Deviating from that order is defensible. Deviating from it without knowing you have deviated is what produces the sixty-person company with an excellent automation stack, three different pipeline numbers, and nobody with the standing to decide which one is real.

Related reading

• What a GTM engineer actually does, including the skills that appear in real adverts and how to assess them.
• What each of these roles pays in the UK and US, and what moves a band.
• What to look for in a first ops hire when there is nobody for them to ask.
• RevOps hire or another AE, for the budget question that sits underneath this one.
• How to hire a RevOps manager, the process end to end.
• Why your RevOps job description attracts the wrong people, which is where most of these role mix-ups become visible.