पाठशाला Pathshala · वृद्धि Vṛddhi, Growth · Lesson 06 · Start

The sales funnel and the conversion rates to expect

A pipeline is a list of hopes until every stage is defined and the rate between stages is measured. The funnel from lead to cash, the rates to expect in India, and how to run it backwards.

Pathshala, The Founder Library · 11 October 2026 · 8 min read

A founder in Noida tells her investors the pipeline is ₹2 crore. It is a spreadsheet of forty companies, eleven of which replied to an email once, and the two that are about to close have been about to close since February. Nobody is lying. There is simply no funnel, only a list.

A funnel is the list with two things added: a definition of each stage that two people would apply the same way, and the measured rate at which prospects move from one stage to the next. With those two things a founder can forecast, can find where deals die, and can answer the question this lesson is built to answer: how many calls this week for the customers wanted this quarter. David Skok’s SaaS Metrics 2.0 supplies the method, Steli Efti and Gong supply the rates at the top, and the Indian specifics supply the stage most founders leave off the end.

Stages you can defend

Skok’s instruction is the whole discipline: for each step in the funnel measure two things, the number that entered it and the rate at which they converted to the next. That only works if the steps are defined, so write them down. Lead: a named person at a company that fits the one-sentence buyer, with a way to reach them. Qualified lead: has the problem, has the authority or a direct line to it, has a reason to act this quarter, and the company can pay the price. Meeting: a conversation of fifteen minutes or more with the decision-maker about their problem; a demo to an intern is not a meeting. Proposal: a price and terms in writing, sent to the person who signs. Closed: a purchase order, a signed quotation or a payment. Verbal agreement is not closed. A16z’s 16 Startup Metrics is specific that letters of intent and verbal agreements count neither as bookings nor as revenue, and the discipline applies at every stage below.

Then the Indian stage: collected. In most Indian companies closing and being paid are different events separated by weeks, and in SMB sales sometimes by a second negotiation. The [runway lesson](/library/runway-how-many-months-you-really-have) is specific that receivables are not cash; the funnel should be too. Track closed-to-collected as a rate and as a number of days, because a business that converts well to orders and poorly to payment has a sales process that works and a company that does not. Efti’s rule for pilots makes the same point from the other side: charge for them, and do not count the money as revenue until the customer has stayed six months.

Two outcomes deserve their own columns. Lost means they chose something else, and the reason goes in the row. No decision means they did nothing, and it is the most common and the least recorded way a deal ends. A funnel that records no-decision honestly will show most founders that their real competitor is inertia, which changes the pitch.

The rates to expect, and where they come from

There are no published conversion benchmarks for Indian B2B sales worth building a plan on, and most of the global ones are vendors’ marketing. What follows are the anchors that can be verified, and beyond them the rules of thumb operators repeat, labelled as such. At the top, cold outreach to a stranger: Gong’s analysis of 90,380 cold calls found a baseline of about 1.5 per cent of calls producing a booked meeting, rising with technique; Efti’s minimums in his Stripe Atlas AMA are about fifteen per cent of cold calls answered and fifteen per cent of those leading to something, and for email a twenty to thirty per cent open rate with ten to twenty per cent of openers replying; Woodpecker’s twenty-million-email dataset puts the average reply rate across all cold email at about 3.4 per cent and good sequences at five to ten. Qualified lead to meeting, where the lead has already shown interest or come through a referral: a quarter to a third is the figure operators repeat, higher for warm introductions. Meeting to proposal: around half, if qualification was honest; much higher means the meetings are being held with people who were already going to buy, much lower means the qualification is not being done. Proposal to closed: a fifth to a third in SMB where the owner decides, and this is the rate most sensitive to price and to how the ask was made. Skok’s one hard number sits beside it: for products with a free trial, trial-to-paid should ideally run at fifteen to twenty per cent.

Enterprise runs lower and slower at every stage below the first meeting, for reasons that have nothing to do with the product: more people in the decision, a procurement function, vendor registration, a security questionnaire, a legal review, and a purchase-order cycle with its own calendar. Lead-to-meeting around a tenth, proposal-to-closed around a fifth or below, and a cycle measured in quarters, with the first deal in any enterprise taking longest. Bain’s India SaaS Report 2021 describes the two motions Indian companies actually run, SMB-focused “India for India” and enterprise-focused “India for the world”, and notes the price gap that sits under the SMB funnel: the same Zoho Books plan listed at about $151 a month in the United States and about $101 in India. Indian SMB buyers decide fast and pay less, and the funnel has to be built for both facts.

Running it backwards

Skok’s planning method is to start from the revenue wanted and work up the funnel to the leads and the people needed. In rupees: a company at ₹60,000 a year per customer wants ₹30 lakh of new annual revenue this quarter, which is fifty customers, or about seventeen a month. At thirty per cent proposal-to-close that is fifty-six proposals a month. At fifty per cent meeting-to-proposal, a hundred and eleven meetings. At twenty-five per cent lead-to-meeting, about four hundred and forty qualified leads a month, which is a hundred a week to find, reach and qualify. A hundred and eleven meetings a month is twenty-six a week. One founder selling full time can hold twenty. The arithmetic has just said two things the founders did not want to hear: the lead list is the bottleneck, and the first sales hire is due before the quarter ends, not after.

Run your own numbers through the figure. The presets are starting points, not benchmarks. After fifty logged conversations from the [hundred-calls process](/library/founder-led-sales-first-hundred-calls) replace every rate with your own, and the figure stops being an illustration and becomes a plan.

A pipeline number is a wish. A funnel is a wish divided by a conversion rate, and the division is where the honesty lives.

Reading where it leaks

Each stage leaks for a different reason and the fix is different. Lead to meeting is low: the list is wrong or the message is, in that order; go back to the buyer sentence and to the [outreach lesson](/library/cold-outreach-that-gets-replies-in-india) before touching the product. Meeting to proposal is low: the meetings are with the wrong person, or the founder is pitching before the buyer has named the problem. Proposal to closed is low: price, terms or the ask; look at the no-decision column, because a proposal that dies in silence usually died of a price that was never discussed on the call, or of an ask that had no date. Closed to collected is slow: the PO, the GSTIN, the advance, the named owner on the buyer’s side, all of which belong in the closing conversation and not after it.

Watch days as well as rates. Skok’s funnel is a count; the version that forecasts adds the average days a deal spends in each stage, because a deal that has sat in proposal for three times the average is not in proposal. It is lost and has not been told. Once a quarter, compute the rates by lead source as well, because Skok’s observation that conversion and cost per lead vary considerably by source is what connects this lesson to the [channel test](/library/finding-first-channel-that-works): a channel that fills the top of the funnel with leads that never reach a meeting is the most expensive channel on the scorecard, whatever its cost per lead.

SMB and enterprise in India: what the funnel has to allow for

For SMB, the funnel is short and the owner is the funnel. Lead to closed can take two weeks if the founder is in the room, and the stages that matter are qualification, because owners will take a meeting out of courtesy and never intend to buy, and collection, because the same owner who decided in a day may take sixty to settle. Build in an advance, a monthly plan where the product allows it, and a reminder cadence that is part of the process rather than an awkward call. Expect price objections at every stage and treat them as data on the segment rather than as instructions to discount.

For enterprise, the funnel needs stages the SMB one does not: vendor registration or empanelment, the security and compliance review, legal, and the PO, each of which can take a month and each of which is a place a deal dies without anyone deciding against you. Add them as stages so that they are measured. Charge for pilots and define their success criteria in writing before they start, because an undefined pilot is a free trial with a longer name. And keep the enterprise funnel separate from the SMB one in the sheet and in the figure, because a blended rate across a ₹60,000 deal and a ₹25 lakh deal describes neither.

The weekly pipeline review

Friday, forty minutes, the founders and whoever sells. Five things, in order. Stage counts: how many in each stage today against last Friday, which shows whether the top is being fed. Rates on a trailing ninety days, not this week’s, because this week’s are noise; write them next to the plan rates in the figure and note the gap. Aged deals: every row that has sat in a stage longer than twice the average gets a decision today, a dated next step or a move to lost or no-decision. Collections: every closed deal not yet paid, with the day count and the name of who is chasing. The forecast: the weighted pipeline, each deal’s value multiplied by its stage’s trailing conversion rate, which is the only forecast number an investor should ever be shown and the one that would have saved the founder in Noida from saying two crore. Then set next week’s number for the top of the funnel, because every rate in this lesson is useless if nothing goes in.

Nothing here is financial or legal advice. The rates are starting points and the sources are below; your own fifty conversations outrank all of them.