Revenue operations consulting: buy or hire

Buy revenue operations consulting when the work is a bounded state change with a defined end: a migration, an implementation, an unpicking of two systems after an acquisition. Hire when the work is a rate rather than a state, which is most of revenue operations. The deciding question is not what the two options cost. It is whether the thing you are buying degrades when nobody is tending it.

The short answers

• The question is decay, not price. A consultant leaves a state behind. An employee holds a rate. If the deliverable is still true in twelve months with nobody maintaining it, buy it. If it drifts within a quarter, you are renting attention and the money belongs in a seat.
• Most revenue operations work fails the decay test. Stage definitions, field hygiene, routing, forecast discipline and territory rules all degrade continuously, because people keep adding fields and closing deals in ways the model did not anticipate. That is not a criticism of consultants. It is what continuous work means.
• A small, genuinely bounded set passes it. CRM migrations, first-time implementations, quota and territory design for a coming year, post-acquisition system consolidation, and a diagnostic you need before you can write a job description.
• The published pricing is all vendor pricing. Agency retainers of $3,000 to $27,000 a month and $200 to $400 an hour, fractional tiers of $2,500 to $15,000 a month against eight to sixty hours and $4,000 to $15,000 by ARR band. None of those is a survey. They are list prices published by firms selling the service.
• The consulting case is strongest before the hire, not instead of it. A four to six week diagnostic that tells you what is actually broken is a good purchase, because the most expensive hiring mistake is recruiting against a symptom.
• Handover is where engagements quietly fail. Not because the work was bad, but because nothing was named. Four things have to be in the statement of work: definitions, access, a named internal owner per rule, and a written account of what was not done.
• Be careful with the market’s own numbers. The most-quoted figures in this category are published without samples or methods, including BCG’s claimed 10 to 20 per cent sales productivity gains and Gartner’s prediction that 75 per cent of the highest-growth companies would deploy a RevOps model by 2025. Neither discloses one.

Four different products share one phrase

“Revenue operations consulting” is sold as a single category and is at least four distinct purchases. Getting the wrong one is the most common way this money is wasted, and it happens before any negotiation about price.

What you are actually buying	Typical shape	What it leaves behind	Does it decay?
Diagnostic	Four to six weeks, interviews and a systems audit, written findings	A prioritised account of what is broken and what it costs	No. A diagnosis is true when it is written
Implementation	Eight to twenty weeks, a defined system built or migrated	A working system, configured	Slowly. The build is durable, the conventions around it are not
Framework design	Four to twelve weeks, definitions, stages, territories, quota model	Documented rules	Fast, unless somebody enforces them
Ongoing operations	Monthly retainer, a fractional lead or a managed service	Attention, for as long as you pay	Entirely. It stops when the invoice stops

The first two are usually good purchases. The third is a good purchase attached to a problem, which is that a framework is only worth what its enforcement is worth. The fourth is a seat with a different contract, and should be compared against a seat.

A firm will happily sell you all four together, and there is nothing improper about that. But you should know which one you are relying on for the return, because they fail differently and they need different things from you.

The decay test

Here is the whole decision, in one question. If everybody involved walked away on the day of handover, how long is the deliverable still true?

Score the work, not the vendor.

The work	Still true after handover	Verdict
CRM migrated from one platform to another	Indefinitely. The records are where they now are	Buy
Two CRMs consolidated after an acquisition	Indefinitely, once the cut-over is done	Buy
A first CRM implemented for a company that had none	Years, in its bones. Conventions drift sooner	Buy
Territory and quota model designed for next year	One planning cycle, by construction	Buy
A written diagnosis of why the forecast is wrong	Permanently. A finding does not expire	Buy
Attribution model rebuilt	Until the channel mix changes materially	Borderline
Pipeline stage definitions rewritten	About a quarter without enforcement	Hire
Field and object hygiene cleaned up	Weeks. People add fields	Hire
Lead routing rules rebuilt	Until the first territory change	Hire
Forecast discipline in the weekly call	Until the next call nobody chairs	Hire
Deciding what a qualified opportunity means, and holding it	It is a rate, not a state	Hire

The pattern is not subtle once you look at it in a column. The things that survive are the ones where the change is physical: records moved, a system built, a number computed and written down. The things that decay are the ones where the change is behavioural, and behaviour reverts to whatever the incentives were before you paid somebody to write a definition down.

This is why the honest version of the consulting pitch is narrower than the sold version. A consultant can genuinely move your data, genuinely build your system and genuinely tell you what is wrong. What a consultant cannot do is be in the room in six weeks when a sales leader wants stage 3 to mean something more convenient.

What the pricing actually looks like, and what it is worth

Every published number in this category comes from a firm that sells the service. There is no independent rate survey for revenue operations consulting, which is worth saying plainly before quoting any of it.

Source	What it publishes	What kind of number this is
MergeYourData	Agency retainers $3,000 to $27,000 a month; hourly $200 to $400	Vendor pricing page, no sample disclosed
OpsEthic	Four fractional tiers, $2,500 to $15,000 a month against 8 to 60 hours	Vendor pricing page, proprietary figures
Mountainise	Fractional tiers $4,000 to $15,000 a month by ARR band	Vendor page, citing two other vendor pages

The IT Jobs Watch figure is the closest thing to a market observation rather than a list price, because it is scraped from live advertisements rather than published by a seller. It is also drawn from eighteen quoted rates, which is a small enough sample that the median should be read as an indication and not a benchmark. That is the state of the evidence in this market.

What none of this pricing tells you is the thing that decides value, which is the ratio of senior hours to junior hours in the delivery. A $15,000 monthly retainer delivered by the person who ran the sales process is a different product from the same retainer delivered by an analyst with the principal on a fortnightly review. Ask for the split in writing.

For what the permanent alternative costs, the UK and US bands are published in full, and the fully loaded arithmetic of an operations seat against a quota-carrying one is worked through in RevOps or another AE. This article is not going to rebuild that model, because the argument here is that cost is the second question and shape is the first.

The engagement structures, and how each one fails

Structure	Best for	The characteristic failure
Fixed-price project	A bounded build with a clear specification	Scope defined tightly enough to price becomes scope defined tightly enough to exclude the hard part
Time and materials	Discovery, or a build with genuine unknowns	No natural end. The engagement becomes a habit
Monthly retainer	Continuous operations you have decided not to hire for	You are paying seat money without seat continuity
Fractional lead	A bridge to a permanent hire, with a date	The bridge becomes the structure, and nobody ever runs the search
Outcome-based	Rare, and usually only for a single measurable number	Attribution arguments. Almost nothing in ops has a clean counterfactual

The fractional row deserves a sentence more, because it is the option most often proposed as a permanent answer. A fractional lead is genuinely good at the planned half of operations work: the monthly reporting, the quarterly territory review, the arbitration you can schedule. Operations is substantially unplanned: the pricing exception, the field somebody added, the deal that does not fit the model. Unplanned work needs somebody who is present when it happens. Fractional arrangements that work have a named end date written into them at the start.

Buy the diagnosis before you buy the hire

The strongest case for consulting is not as an alternative to hiring. It is as the thing that happens first.

The most expensive hiring mistake in this category is recruiting against a symptom. “The forecast is wrong” is a symptom with at least five causes: a stage model nobody follows, a data layer that disagrees with itself, a sales leader who commits optimistically, a product with a retention problem showing up as churn in the wrong quarter, or a pipeline too thin to forecast at all. Each of those needs a different hire, and two of them need no hire whatsoever. Companies that skip this step tend to hire a capable person into an undefined mandate, which is the single most common reason ops hires fail inside ninety days.

A four to six week diagnostic, bought deliberately as a diagnostic, produces the thing you need in order to write a scorecard: a specific, prioritised, costed account of what is broken. That output does not decay. It is also cheap relative to a mis-hire.

Two conditions make this work. Buy the diagnosis from somebody who is not bidding for the implementation, or accept that you are reading a proposal. And insist that the output names what is not worth fixing, because a diagnostic that finds everything important has found nothing.

If the diagnosis says the problem is continuous, you now have a scorecard and a defensible order of steps for the search. If it says the problem is a bounded state change, you have a specification, and you can buy that state change from anyone.

The handover problem

Engagements rarely fail on the quality of the work. They fail at the point where the work becomes somebody else’s, and they fail quietly, over about two quarters, in a way that is easy to mistake for the consulting having been useless.

Four things prevent most of it, and all four belong in the statement of work rather than in a conversation near the end.

Definitions, written where the team already looks. A stage model that lives in a PDF in a shared drive is not a stage model. It has to be in the tool, in the field help text, in the onboarding, in the place where somebody about to make the wrong choice would encounter it.

Administrative access held by your own staff throughout. Not transferred at the end. Held from the start. If the consultant’s account is the only one with the permissions to change what they built, the handover has a dependency in it that nobody has priced.

A named internal owner for every rule that changed. Not a team. A person, who knows they own it, and who was in the room when the decision was made. Rules without owners revert.

A written account of what was deliberately not done. This is the most-skipped and the most valuable. Six months later, the question “did anyone consider X?” has an answer, and the person who inherits the system can tell the difference between a decision and an oversight.

Write these in before signing. A vendor who resists all four is telling you something useful at no cost.

When consulting is clearly the right purchase

• A migration or consolidation with a cut-over date. Bounded, physical, and needing a skill you will genuinely not need again for years.
• You need the diagnosis before the job description. Covered above, and the most under-bought version of this.
• A one-off design task on an annual cycle. Territory and quota design for the coming year is a project with a natural end, and buying it does not stop you hiring later.
• A specialist skill for a specific system. Deep platform work on a system nobody in-house knows is a reasonable thing to rent.
• You cannot fund an eighteen-month seat. If the runway will not carry the role, buying a bounded improvement is a better decision than hiring somebody you will have to let go, and it is a more honest one toward the person you would have hired.
• The permanent search is running and something is on fire now. Bridge cover with a date is a legitimate use, and it is the case the fractional model is genuinely built for.

When hiring is clearly the right decision

• Somebody has to arbitrate. When sales, marketing and finance produce three numbers, the job is to decide which is correct and make that stick. Authority of that kind does not attach well to a contract. Forrester found centralised revenue operations rose from 15 per cent to 40 per cent across two years among 316 respondents, which measures centralisation rather than reporting lines, and centralising a function into a contractor is a contradiction.
• The work is continuous and you know it. Everything in the “hire” column of the decay table.
• The unplanned half matters. If the cost of a four-day wait for an answer is material, you need presence, not availability.
• You are at the size where the ratio says so. ICONIQ’s 2025 survey of 205 go-to-market executives put revenue operations at 6 per cent of go-to-market headcount, 7 per cent at high-growth companies. Practitioner heuristics put the first generalist at around twenty-five go-to-market staff, then one per thirty to forty, and a rule of thumb circulates from named practitioners with no sample behind it. Treat the survey figure as budgetable and the heuristics as a sanity check, because that is what their own authors call them.
• Institutional memory is the point. Knowing why a field exists is worth more than knowing what it does, and that knowledge only accrues to somebody who stays.

Note that the survey figures above are drawn from B2B software companies, most of them American. If you are hiring into the UK, the ratio is a starting point rather than a local benchmark.

One legal point, for UK buyers

If you are engaging an individual consultant through their own limited company in the UK, the off-payroll working rules determine whether they should be taxed as an employee for that engagement, and for medium and large private-sector clients the responsibility for making that determination sits with the client rather than the contractor. GOV.UK sets out how the rules work. This matters to the buy-or-hire decision in a specific way: an engagement structured to look like a seat, one that is sustained, directed, exclusive and indefinite, is exactly the shape most likely to fall inside the rules, which removes part of the flexibility that made buying attractive. Take advice on your own facts. This article is not it.

What the published evidence does not support

The claims made for this category are considerably stronger than the evidence behind them, and a buyer should know which is which.

The most-cited outcome figures do not disclose their methods. BCG’s paper on go-to-market operations claims 100 to 200 per cent increases in digital marketing return, 10 to 20 per cent increases in sales productivity and 30 per cent reductions in go-to-market expense, published with no sample, no method and no measurement period. Gartner’s 2021 prediction that 75 per cent of the highest-growth companies would deploy a RevOps model by 2025 is a prediction rather than a measurement, and is published without a sample or a stated methodology. Both are quoted constantly in this market, usually without those qualifications. Neither should carry weight in a purchase decision.

There is no published study we are aware of that measures the return on revenue operations consulting against a control. The pricing is vendor pricing. The staffing ratios are a single survey plus a set of heuristics that identify themselves as heuristics. That is genuinely the state of it, and anybody presenting this category as a solved arithmetic problem is selling.

What that means practically is that the decay test is more useful than any benchmark you will be shown, because it depends only on facts about your own work rather than on somebody else’s disclosed-free number.

The shape of the work decides, not the invoice

Ask what happens to the deliverable when everyone walks away. If it is still true in a year, buy it, and buy it from whoever is best at that specific bounded thing. If it drifts within a quarter, no engagement structure will fix that, because you are trying to purchase a rate with a project.

The sequence that works for most companies is not one or the other. Buy a diagnosis, use it to write a scorecard, hire against the scorecard, and buy bounded projects afterwards for the state changes that come up. That order gets the consulting spend into the part of the problem it is genuinely good at, and puts the continuous work where continuous work belongs.

Related reading

• What to look for in a first ops hire, if you have not hired any operations person yet and the readiness question comes before this one.
• RevOps or another AE, for the fully loaded cost arithmetic this page deliberately does not repeat.
• How to hire a RevOps manager, for the process once the diagnosis says the work is continuous.
• Why ops hires fail in the first 90 days, for what an undefined mandate costs after handover.
• GTM engineer versus RevOps manager, for which operations role the diagnosis actually points at.
• RevOps salary benchmarks, for the UK and US bands behind the permanent alternative.
• Recruitment fees explained, for what the hiring route costs alongside the consulting route.