पाठशाला Pathshala · वृद्धि Vṛddhi, Growth · Lesson 30 · Scale

Revenue operations: the function that keeps growth honest

When marketing, sales and customer success each keep their own numbers, nobody can say what the quarter will bring. Revenue operations builds one record, one definition of every stage and a forecast whose errors are measured.

Pathshala, The Founder Library · 11 October 2026 · 7 min read

A railway signal gantry with lit indicators against a dusk sky. — Photograph: Jean-Francois Frenel · Pexels

The board asks what the quarter will close at. The head of sales says ₹6 crore, marketing’s dashboard implies ₹8 crore, finance has booked ₹4 crore and the CRM, if anyone opened it, would say ₹11 crore. Every number is defended by someone. None of them is the forecast.

Companies grow into this. In the first year the founders know every deal. By the time there are a few sellers, a marketing team and someone looking after customers, each function keeps its own spreadsheet and its own definition of a lead, an opportunity and a customer, and the numbers stop adding up. Revenue operations, RevOps, is the function whose job is to make them add up: one record, one set of definitions, one process from first touch to renewal, and a forecast that is measured against what actually happened. This lesson sets out what it does, when to start it and the calendar it runs on.

What revenue operations is

RevOps owns four things across marketing, sales and customer success. Data: the CRM and everything that feeds it, its fields, its integrity and its reports. Definitions: what counts as a lead, a qualified opportunity, each stage, a booking, a churned customer, written down once and used by everyone. Process: how a lead is routed, how a deal is handed from seller to customer success, how quotes are approved, how territories and quotas are set. Tools: which systems the revenue teams use, and which they stop using. It does not own the targets or sell; it makes the numbers that targets are judged against trustworthy.

The need is not unique to Indian startups. Gartner’s survey of sales operations leaders, published in February 2020, found that only 45 per cent of sales leaders and sellers had high confidence in their organisation’s forecasting accuracy, and only 47 per cent believed their organisations had high-quality data. And Salesforce’s State of Sales research, from 7,775 sales professionals in 38 countries surveyed in August and September 2022, found reps spending just 28 per cent of their week actually selling, using an average of ten tools to close deals; 65 per cent of sales leaders described sales operations as key players in defining strategy, up from 54 per cent in 2020.

Start the function when the company has five or more people selling, a board that expects a quarterly forecast, or more than one team touching the same customer. Before that, a founder who owns the CRM and the [weekly metrics review](/library/weekly-metrics-review-one-page-one-hour) is enough.

One definition of every stage

Every pipeline report rests on stage definitions, and most are written as activities: demo done, proposal sent. Write them instead as what must be true to leave the stage, as the [sales playbook](/library/sales-playbook-from-heroics-to-repeatability) does. A lead becomes a qualified opportunity when discovery has confirmed the problem, the budget owner is known and there is a reason to act this quarter. An opportunity reaches proposal when the buyer has agreed the scope. It reaches commit when the buyer has confirmed the decision process, the paper process and the date. A deal is won when the contract is signed, not when the buyer says yes on a call.

Define the edges too: what makes a lead marketing-qualified and who must accept it; when an opportunity is closed-lost, with a required reason; when a customer counts as churned, and whether a downgrade is churn. Put the definitions in one document, link it from the CRM, and change it only at a quarterly review with every function in the room. When marketing and sales disagree about numbers, the disagreement is almost always about a definition.

Pipeline hygiene

A pipeline is honest only if dead deals are closed and live deals are current. Pete Kazanjy’s Founding Sales puts the rule bluntly: if it is not in Salesforce, it does not count, and he recommends error-checking reports that flag open opportunities with too little activity or tasks not executed, a weekly one-hour pipeline meeting and closing opportunities that will not happen so time goes to those that will. Whatever CRM you use, the rule is the same: meetings, emails and decisions that are not in the record did not happen as far as the forecast is concerned.

Curved railway tracks and signals at an urban station. — Tracks that are not maintained still look like tracks from a distance. Pipeline that is not cleaned still looks like pipeline. Photograph: Ayşe · Pexels

Make the hygiene mechanical. Every Monday, a report lists open opportunities with a close date in the past, no next step, no activity in fourteen days, or a stage change without its exit criteria. Each owner fixes or closes them before the pipeline meeting. Track the share of pipeline that is stale over time; when it is falling, the forecast is getting more honest. Resist the opposite failure, a CRM with sixty mandatory fields; every field a seller must fill is time not spent selling, and fields nobody reads should be removed.

Hygiene runs past the signature. A won deal is handed to customer success with the scope, the stakeholders and the promises the seller made, recorded in the CRM rather than in a call. Renewals and expansions go into the same pipeline with the same stages and the same rules, because a renewal that slips is as real a miss as a new deal that does, and in a subscription business it is often the larger number. Customer success owns those opportunities and forecasts them in the same categories. A company that forecasts new business carefully and renewals by assumption is forecasting half its revenue.

A forecast a board can trust

A weighted pipeline, each deal’s value times a probability by stage, is a useful check and a poor forecast, because stage probabilities hide what the seller knows about each deal. Forecast instead by category. Commit is the deals the owner will be held to, each with a confirmed decision date and paper process. Best case is the deals that could close this period if things go well. Pipeline is everything else qualified. The forecast for the period is commit plus a judgement on best case, made by the sales head deal by deal, and the weighted pipeline sits beside it as a cross-check. The [annual plan lesson](/library/forecasting-and-annual-operating-plan) covers how the forecast feeds the plan.

Then measure the forecast. At a fixed point, say the end of week two of each month or quarter, record the forecast. At the end, record the actual. The error, forecast minus actual divided by actual, and its direction, by rep and in total, tell you more than any model: a rep who over-forecasts by 30 per cent every quarter is giving you information, and so is a team that always lands 10 per cent under. After four quarters, show the board the record. A forecast whose past errors are visible is trusted more than one presented with confidence.

A forecast is trusted not because it is confident but because its past mistakes are written down.

Fewer tools, one record

The Salesforce research found two-thirds of reps overwhelmed by the number of tools and 94 per cent of sales organisations planning to consolidate their stacks. RevOps owns that consolidation. Keep the CRM as the only record of accounts, contacts, opportunities and activities, and connect every other tool to it, the email and calendar, the calling tool, the marketing system, billing and the support desk, so that nothing has to be typed twice. In India, connect the billing or accounting system too, so that a booking in the CRM can be matched to an invoice and a GST return; the [monthly close](/library/monthly-close-and-mis-report) depends on it. When the company outgrows the CRM’s reports, the [data warehouse lesson](/library/first-data-warehouse-single-source-of-truth) is the next step.

Who runs it and where it reports

The first RevOps hire is usually an analyst with sales operations experience who can administer the CRM, build reports and is comfortable telling a senior seller that a deal does not meet the criteria for commit. Report the function to whoever owns revenue, often the chief revenue officer or the founder who runs sales, with a dotted line to the finance head so that bookings and revenue reconcile. Do not report it to the sales head alone if that person is also judged on the forecast; the function exists to be independent of the number.

As the company grows, the function adds compensation administration, territory and quota planning, deal desk for pricing approvals, and enablement for new sellers. Add them in that order of pain. The checklist below is the calendar the function runs on from the first hire.

The RevOps calendar

Weekly, the function keeps the pipeline clean and the forecast current. Monthly, it reconciles with finance and measures the forecast. Quarterly, it revisits definitions and reports to the board. Tick it off as you go.

The survey figures are from the sources below and describe large samples, not your company. Write the stage definitions this week.