What a modern revenue operations organization looks like and how the work is divided between people and machines.
Most Revenue Operations (RevOps) functions in place today were designed five or six years ago, when the job was to get sales, marketing, and customer success onto one set of systems and one set of numbers. Today, RevOps’ goals remain primarily unchanged: deliver accelerated revenue at scale, efficiently. What has changed is that AI now does a real share of the work: scoring, monitoring, routing, summarizing, and increasingly recommending what to do next.
A function built to consolidate data and processes isn’t automatically built to direct machines, and that is where many teams get stuck. RevOps teams add AI tools to an organization that was never designed to decide which work a machine should do, which a person should do, and who is accountable when the two disagree.
This guide explains how to build and run a modern RevOps organization that leverages AI independently and jointly with humans. It covers four key areas:
What RevOps is for, and why a clean, integrated foundation matters more once AI is acting on the data
The three modes of work, and how responsibilities divide between AI and people
What the function owns: six capability areas held together by three layers that span all of them
How the function is structured, how it runs through a year, and how to read where your own organization stands
The instinct is to treat RevOps as a cost to be managed, and that framing undersells it. When people, process, and technology genuinely align across revenue teams, the function stops being support overhead and starts driving growth, productivity, and efficiency. The published research points in one direction.
Figures compiled from published analyst and advisory research, reported directionally. Gartner has separately projected that a large majority of the highest growth companies will run a RevOps model.
Forrester
More revenue growth for companies that align people, process, and technology across revenue teams.
Forrester
Higher internal productivity from the more mature RevOps teams.
Boston Consulting Group
Increase in sales productivity for firms that leaned on RevOps to accelerate growth.
Boston Consulting Group
Reduction in go-to-market expenses reported by companies that invest in RevOps.
A function that owns a shared data model, one set of definitions, and a standard the geographies actually follow produces those outcomes. A function that runs reports on request does not, no matter how many people it has.
The practical consequence is a different funding conversation. A cost center defends its headcount. A lever argues for investment against a return, so the function must state what it will move and by how much before the budget cycle, not after.
Integration was already the direction of travel before AI. Companies were consolidating systems, aligning definitions, and bringing departmental operations teams into one function because siloed operations created a fragmented view of the customer, friction at every handoff, overlapping tools, and metrics that pitted teams against each other.
AI raises the stakes. When a person works from fragmented data, they notice the inconsistency and compensate for it. When a model works from fragmented data, it scores, forecasts, and recommends at full speed on top of it, and every decision downstream inherits the error. One governed source of truth stops being a nice-to-have and becomes the precondition for any AI-driven revenue operations model to work.
The same logic applies to speed. Much of RevOps’ value shows up in how quickly a revenue signal becomes a decision, then an action a seller can take. When that loop is fast, the forecast updates while the quarter is still open and guidance reaches the rep while the deal is live. When it is slow, AI simply produces stale answers faster. A clean, connected foundation is what makes the loop fast.
A modern RevOps organization assigns every responsibility to one of three modes when it designs the role. The split follows what each side does best. AI is faster and more consistent at high-volume, pattern-heavy work. People are stronger where context, relationships, and accountability decide the outcome. In between sits a large amount of work where a person and a machine on the same task do better than either would alone.
AI-led. Software does the work and a person spot-checks it. The task is repetitive, the volume is high, and catching a wrong answer is cheap. Examples: reviewing every sales call transcript against the ideal customer profile and the agreed sales process, flagging deals whose qualification fields contradict the call notes, deduplicating and enriching CRM records, and tracking whether reps are actually using the playbook.
Co-led. AI assembles the picture, and a person makes the call. The analysis is heavy, but the decision requires judgment. Examples: a weekly deal review where AI ranks risk across the pipeline and the manager decides which deals get attention; a coaching plan where AI identifies skill gaps and the leader picks the priorities; and a territory rebalance where AI models the options and RevOps chooses one.
Human-led. A person owns the work and uses AI to stay informed. The outcome depends on trust, relationships, or accountability that cannot be delegated. Examples: discovery conversations with a buyer, building the business case with a customer’s CFO, negotiating a renewal, and choosing the sales methodology the whole team will run.
The line between the modes moves. Tasks that need a person in the loop today will move to AI-led work as the technology improves, and that movement is where the value comes from. Some of it is efficiency, doing existing work for less. More of it is productivity, because operational work that used to be scarce becomes available to every team in the revenue model.
Because the line moves, someone has to own where it sits. That is governance, and it is the part most organizations leave implicit until something goes wrong: an AI-generated forecast nobody signed off on, a lead-routing model that quietly changed who gets what, a next-best-action recommendation the field stopped trusting. In a modern function, every consequential machine decision has a human owner, and the rules for when a task can move from one mode to another are written down.
The RevOps charter holds those rules. The framework in the next sections describes the work. The charter authorizes it, and when questions come up about whether the work is right or what should happen next, the function and its stakeholders return to the charter. It sets out what the function owns, what decisions it can make on its own, where its scope meets Finance, Sales, Marketing, and Customer Success, and how it settles disagreements. A new AI capability, a reorganization, an argument about whose number is correct: each of these is a question the charter should already answer.
It reads more like a decision-rights agreement than a mission statement, and a RevOps leader takes it to their executive sponsor for written agreement. It has seven components: a North Star, scope and boundaries, decision rights, roles and ownership, success measures, interfaces with each partner function, and the operating rhythm.
That’s why we say organizational design starts with people. Every capability described below is built and run by people, and the first design decision is ownership. Tools come after.
The three modes describe how work gets done. The operating model describes what RevOps delivers. It has six capability areas and three layers that span them all, and together they are accountable for a defined set of revenue outcomes: pipeline quality, win rate, deal size, cycle length, retention, and forecast accuracy.
A useful way to keep the two in mind is to think of a house. The six capability areas are the rooms, each with a distinct purpose. The three layers are the wiring, the plumbing, and the foundation that run through every room. You do not live in them, but nothing in the house works without them.
Revenue and financial planning
This area turns the revenue target into an operational plan and keeps the cost of hitting it in view. Finance and FP&A own the top-line forecast. This area owns the reality beneath it: capacity, coverage, pipeline sufficiency, and the scenario work behind resourcing tradeoffs. A typical problem it solves: the plan calls for 30 percent growth next year, and this area works out how many ramped reps that requires by quarter, whether talent acquisition can deliver them in time, and what the number looks like if it cannot, before Finance locks the plan.
Technology and infrastructure
This is the connected platform everything else runs on: the CRM and systems of record, the integrations between them, the data environment, the tool roadmap, and security and access. For a non-technical reader, the test is simple. Can you add a new tool, or retire an old one, without a six-month integration project? A clean stack makes that routine. A tangled one taxes every initiative built on top of it. A typical problem it solves: three teams bought three call-recording tools, none of which write back to the CRM, and this area rationalizes them to one that does.
Intelligence and orchestration
This area decides where the combined effort of people and machines goes next. It turns data and process into direction: predictive scoring, next best action and guided selling, revenue intelligence on deals and calls, capacity orchestration, and performance measurement. A typical problem it solves: pipeline coverage looks healthy at three times the target, but the model shows half of it is stalled, so effort is redirected to the deals that can still close this quarter. Its output is only as good as the data beneath it, which is why it sits next to the data foundation.
Data and analytics
This is the foundation for everything else in RevOps, and the area that gives the rest of it meaning. It defines what the data means, consistently, everywhere it is used: the single source of truth, shared definitions, data quality and governance, reporting and self-service analytics, and enrichment. When this area is strong, you can trust every model, report, and decision above it. When it is weak, meetings are spent arguing about whose number is right. A typical problem it solves: Sales, Marketing, and Finance each report a different pipeline figure, and this area establishes one definition of a qualified opportunity and one place it lives.
Process and workflow
This is where strategy becomes daily execution, and it is how teams experience working together. Lifecycle design, cross-functional handoffs, SLAs, territory and quota alignment, deal desk, and quote-to-cash all live here. A typical problem it solves: marketing-qualified leads sit for four days before a rep touches them, and this area sets the handoff standard, the SLA, and the routing that gets that down to four hours.
People and organizational design
This area decides who runs all of it: the operating model and team structure, roles and job families, a competency model for the AI era, workforce and skills capacity, and adoption and change management. It turns a framework on paper into an operating organization, whether that is one person covering several areas or dedicated teams for each. A typical problem it solves: the function has hired analysts who can build dashboards, but no one can evaluate whether a model’s output should be trusted, and this area redefines those roles.
How the areas reinforce each other
The six aren’t independent teams, and how they compound makes this an operating model rather than an org chart. Follow one definition through all of them. Data and analytics decide what a qualified opportunity means. Intelligence and orchestration score against that definition, so its predictions are only as trustworthy as the definition beneath them. Process and workflow routes on that score and writes the outcome back, which is what gives data and analytics the evidence to sharpen the definition next quarter.
Technology and infrastructure carry all three, so a tangled stack slows each one at once. Revenue and financial planning consumes the result as coverage and capacity, and it can only plan as accurately as the pipeline view it inherits. People and organizational design decide who owns each link in that chain, and that decision determines whether the chain holds at all.
That pattern runs throughout the model. Strengthen an area and the ones above it get better inputs. Weaken one, and the ones above it inherit the weakness rather than compensating for it, which is why the highest-leverage investment is usually further down than where the symptom appears.
The three layers
AI, governance and ethics, and always-on transformation are layers, not areas, because each runs through every capability area.
AI is a layer because the same underlying capability powers scoring in intelligence, automation in process, and enrichment in data. The layer carries the value framework that keeps adoption deliberate, the operations behind models and agents, and the literacy to direct and judge AI, which is a different skill from using it.
Governance and ethics keep a human owner attached to every consequential machine decision. It runs at three levels, strategic, operational, and technical, holds one standard across all geographies, and is anchored in the charter described above. It manages the risk of autonomy scaling ahead of accountability.
Always-on transformation treats commercial transformation as continuous, because each new class of AI reopens the work. It carries the business readiness assessment, maturity loop, operating rhythm, and change capability that make each new way of working stick.
Three ways to organize RevOps
Centralized. One team sets shared process, technology, and metrics for everyone. This fits a homogeneous business: a company that sells one product line through one motion, such as a mid-market software company with a single sales team and a single kind of customer.
Decentralized. RevOps teams sit inside each segment or region, close to the work. This fits businesses whose units sell in fundamentally different ways, such as a company with an enterprise software business and a separate advertising business, where the buyers, sales cycles, and metrics have little in common.
Hybrid. Local teams own work close to the business, and one horizontal team, agnostic to segment and geography, owns the shared foundation. This fits most companies past a certain scale, and it is the model this guide builds out.
Two principles hold across all three. RevOps sits outside any single function, with executive sponsorship and the authority to drive change across departments.
How RevOps grows with the business
As the business grows, RevOps grows with it because it becomes more complex, and more decisions depend on data only a coordinated function can keep reliable. Under $10M in annual revenue, one or two people keep the CRM clean and define the customer stages. Between $10M and $50M, a dedicated team builds the single source of truth. From $50M to $100M, the focus shifts to governance and forecasting. Between $100M and $250M, AI enters workflows, and the team shifts to a hybrid structure. Above $250M, RevOps runs globally across every unit. Each stage builds on the last, and doing it this way avoids expensive rebuilds later.
The hybrid organization
The hybrid model has two halves. Local teams, organized by region or segment, sit alongside the sales, marketing, and customer success groups in their market and support day-to-day execution: tailoring customer journeys, creating relevant content, refining local processes, and driving adoption on the ground. Each local team has an operational leader who acts like a chief operating officer for that unit. Whatever the title, the job is the same: keep the local business running, adapt the shared standards to the market, and feed what they learn back to the center.
The horizontal team keeps the whole system unified. It owns the shared data model, systems, core processes, and quality standards every local team depends on. It is organized by capability rather than by region for a simple reason: the things it owns only work if there is one owner for each. Four core groups run the foundation: data and analytics, systems and tools, foundational process, and reporting and insight. Two more sit on the transformation side. The AI Center of Excellence is both a think tank and a deployed resource for AI adoption. Strategic Business Operations is the internal consulting group that runs the major initiatives coming from the business units and the C-suite.
Running the business and building what comes next
Every RevOps team splits its time between keeping the business running and building for the future, and the balance depends on the team. The four core groups spend roughly 80 percent of their time running the business, refining forecasts, keeping data healthy, prioritizing pipeline, and clearing quote-to-cash roadblocks, and 20 percent on initiatives. The AI Center of Excellence splits evenly. Strategic Business Operations sits mostly on the build side, leading the transformative projects for leadership.
Both sides run all year. The annual planning cycle moves through the calendar, building next year while the current one is delivered. The calendar below assumes a January to December fiscal year. If yours runs differently, the rhythm is the same, so map the four phases to your own quarters.
First quarter: execute the plan. Territories, quotas, and comp plans go live; new hires are onboarded and ramped against coverage; and the year’s reporting baseline is built. Track attainment against the new plan, and keep the first forecast honest. Revenue and financial planning and process and workflow lead, with data and analytics standing up the baseline and Finance, Sales, and enablement partnering.
Second quarter: mid-year checkpoint. The forecast is rebuilt on first-quarter actuals, coverage is adjusted and territories reassigned where needed, and scoring and forecasting models are tuned against real results. A mid-year diagnostic reads the capability profile to catch drift before it compounds.
Third quarter: assess and design. The planning cycle opens. The organization runs the full maturity reassessment across all six areas, then aligns leadership on next year’s three to five core goals, the joint OKRs, and the baseline definitions. The AI Center of Excellence and Strategic Business Operations lead, with the C-suite and unit leaders owning the goals.
Fourth quarter: plan and lock. The business runs to close the year while planning peaks. Teams run discovery, diagnostics, and the business case; prioritize and cost the initiative set; and design next year’s territories, quotas, and comp, with executive leadership, the board, and Finance approving at a final decision gate. The quarter ends with funded priorities, balanced territories and quotas, and an agreed plan.
Everything above describes the target. A maturity profile tells you how far from it you are, and it is the most useful artifact a RevOps leader has for making the case to invest, because it replaces “we need more RevOps” with “our data foundation is at level one, our intelligence layer is at level three, and here is what that mismatch is costing us.” It works best when the six area owners and the leaders of sales, marketing, and customer success score it together, using evidence from the systems rather than what each team hopes it is doing.
Each capability area is scored on a four-level scale.
Fragmented means work relies on individual effort and results are inconsistent. Structured brings shared standards and useful automation, with solid wins locally.
Scaled means teams operate on a unified, trusted foundation everyone can act from. Adaptive means the organization measures continuously, absorbs new methods smoothly, and stays ahead of change. Adaptive is not a permanent finish line. As the technology changes, staying there takes ongoing work.
Targets vary by scale. Under $50M, aim for Structured across the core operations. From $50M to $250M, work toward Scaled. Above $250M, aim for Adaptive in the areas that drive the most revenue. Scoring all six areas across marketing, sales, and customer success gives a complete picture of the function’s health.
Where organizations lose effort
Most wasted effort traces to two mismatches. A vertical gap is one area running well ahead of the foundation it depends on, for example, a sophisticated scoring model sitting on unreliable data, where every extra hour produces noise instead of results. A horizontal seam is the same capability at very different maturity levels in different teams or regions, where work stalls at the handoff.
To fix a vertical gap, raise the lower area to match, rather than adding to the higher one. The fix for a seam is to bring the trailing side up to the one ahead of it. The sequence is five moves, run in order and repeated: locate the widest gaps and worst seams, diagnose the real constraint beneath the top one, strengthen it with a named owner and a date, instrument the outcome against a baseline, and reassess after two quarters.
Use this to see the whole model at a glance and mark what you already have in place.
The six capability areas
Revenue and financial planning
Technology and infrastructure
Intelligence and orchestration
Data and analytics
Process and workflow
People and organizational design
The three spanning layers
AI
Governance and ethics
Always-on transformation
The structure
A defined operating model: centralized, decentralized, or hybrid
Executive sponsorship and authority to drive change
A horizontal team for the shared foundation
Local teams for execution, with an operational leader per unit
The artifacts
One governed source of truth and shared definitions
A charter naming ownership and decision rights
Joint OKRs with named C-suite owners
A maturity profile scored across the functions
A standing operating cadence
You do not need to build everything at once. Three moves get the model moving in your first quarter.
Run a maturity profile. Gather the six area owners and the function leads for a half-day workshop, score all six areas across the three functions on evidence, mark the target level for your scale, and capture the two or three widest mismatches.
Draft the charter. State the scope, set the decision rights, name the North Star and the success measures, map the interfaces, and take it to your executive sponsor for written agreement.
Close the widest gap. Pick the one mismatch touching the most revenue, find the real constraint beneath it, assign a named owner and a date, instrument the outcome, and rerun the profile after two quarters.
Score the organization before you buy anything. When a team wants to improve a capability, the instinct is to add to the strongest area, because that is where the momentum already is. A maturity profile usually points the other way. The highest-leverage move is often to strengthen a weak foundation beneath an overbuilt layer, or close the one seam that loses the most capacity. Adding a tool before you understand the mismatch pours more capacity into a system that cannot yet convert it.
This is the AI-Driven Revenue Operations guide. It builds on our Revenue Operations guide, which lays out the integrated engine of aligned functions, shared data, and connected processes. This piece describes what the function owns and how it runs once AI is doing a real share of the work.
It also connects to how a RevOps strategy is optimized in practice, covered in Optimizing Your RevOps Strategy, and to the automation that supports it, covered in RevOps Automation Tools.
Modern revenue operations now has three kinds of owners: AI, people, and both. Everything else in the model follows from that.
A clean, integrated foundation, so machines act on trusted data. A charter, so a person is accountable for every consequential decision and for where the line between modes sits. Six capability areas and three layers that describe what the function delivers. A structure and a planning rhythm that turn company goals into owned, funded work. And a maturity profile that shows where to invest next, rerun as each new class of AI reopens part of the work.
The organizations that get the most from the model treat it as a loop they move through deliberately, year after year.
Work with us
Talk to our team about your revenue operations.
This article covers the whole model at a summary level. The posts that follow take each part in depth, starting with the case for funding RevOps as a growth lever rather than as overhead.
Sources
Figures referenced on revenue growth and productivity are compiled from published Forrester and Boston Consulting Group research and reported directionally at the framework level. The use cases, maturity levels, joint OKR example, and time splits represent the RevEng position and are illustrative. Visuals are drawn from the RevEng AI-Driven Revenue Operations guide.