Cropin’s $2.4M Run Rate and the $4.5M It Promised in Two Months

Nathan Latka did the arithmetic out loud. Krishna Kumar had just told him Cropin was working with 185 companies, and that international customers paid an average of about $50,000 a year. “Now can I take 185 customers multiplied by that 50,000 ACV — that puts you guys at about 9, 10 million bucks in run rate right now, is that about right?” It was not. In January 2019, Cropin was doing $2.4 million.

The distance between those two numbers is the most useful thing on the tape. Kumar’s Bangalore agritech company was selling one product into two economies at once, at prices that differed by roughly four times, and the logo count told you almost nothing about the revenue.

Two price worlds, one product

Asked what a customer pays on average, Kumar split the answer in half without hesitating. International accounts averaged “fifty thousand dollars,” with some at $100K and $200K. Indian businesses were “somewhere around 13,000” a year. Same platform, same crop models, a four-fold spread in annual contract value depending on which half of the customer base you looked at.

Nathan kept pressing the arithmetic: at $2.4 million across 185 logos, the blended average was closer to $13,000 — about a thousand dollars a month per customer. Kumar’s answer was that the 185 counted pilots, some signed at $10,000 and some under $30,000 or $40,000, which had not scaled yet. “You can call it as a 4.5 million ARR we are targeting,” he said, “because some customers will be on the pilot stage and then they scale up.”

Billed on managers, not farmers

The bio Cropin supplied to the show carries the numbers that make an agritech deck: 3.2 million acres of farmland digitised, nearly 2.1 million farmers, a footprint in 29 countries. None of those are billing units. When Nathan asked what a “user” actually is at McCain, Kumar said it is a farm manager — someone with maybe 100 farmers in his portfolio. Cropin priced the managers.

2.1Mfarmers reached, per Cropin’s own bio — not one of them a paying user

That decoupling is how a company touches millions of people and still runs at $200,000 a month. It also settles who writes the cheque. McCain grows potatoes for its chips and engages farmers in the background to do it; McCain pays, and uses the platform to manage “their 100,000 growers in different countries.” Three buyer types were paying by 2019:

• Farm-production companies — McCain, ITC, Philip Morris, Syngenta, Sahyadri Farms: businesses that need contracted growers to produce more from fewer inputs.
• Input companies — chemical and fertiliser sellers using the platform for farmer advisory and for managing their retailers and dealers.
• Banks, NBFCs and insurers — underwriting farm loans against Cropin’s data. Close to 10% of the trailing twelve months’ revenue, and the newest of the three.

The lending leg is the one Kumar built the company for. He was a hardcore R&D engineer at General Electric who went looking for “a sector which really needs technology where nobody wants to invest their time.” Indian agriculture came first on the list, he said, because farmers were taking their own lives over debts of around a thousand dollars. The pitch to a bank is that three years of a farmer’s cultivation history opens the black box — it shows the lender this has been a good farmer, and it drops the cost of monitoring and underwriting far enough that agricultural lending becomes a viable business at all.

Nine years to $2.4 million

• 2010 · Founded Kumar leaves GE to build a platform connecting the agri ecosystem, well before agritech was a category in India.
• Late 2011 into 2012 · First product The first working product ships and Cropin starts selling to customers.
• 2015 · Critical mass Close to 88 terabytes of farm data have accumulated; the company starts building machine-learning models on top of it instead of just recording.
• Mar 2018 · Fiscal 2018 closes $1.5 million run rate — north of $100,000 a month.
• Dec 2018 · Series B $8 million raised, taking total funding to $12 million.

Kumar was specific about what the Series B was for. One half was feet on the ground in international markets, because enterprise clients abroad require them. The other was AI and ML: models that predict which plots in Nigeria are growing which crop, what the health of each is, and what the yield will be — then sell that read onward to supply-chain managers, commodity traders, banks and insurers.

The two-month year

Everything closes before March 31

Cropin’s fiscal year ends March 31. At taping, Kumar was at $2.4 million and forecasting $4.5 million by the close — nearly doubling the company inside a single quarter. Nathan did not let it pass: “doubling your company in two months is basically what you’re telling me you’re gonna do.”

All the hard work of a few months are going to close in this one.

Krishna Kumar, founder & CEO, Cropin

Bookings were heavily weighted to February and March every year — “that has been always a heavy year for us in terms of the ARR.” Behind the forecast sat two or three enterprise deals with ticket sizes north of a million dollars, four or five months into negotiation. Nathan’s objection was the one any buyer would raise: if your average customer pays $12,000 a year, who are the references for a million-dollar contract? Kumar’s answer was that customers paying $300,000 a year already existed, and that Syngenta and Philip Morris were live, multi-country and visible to every prospect in the sector.

Expansion measured in countries

Cropin’s land-and-expand did not run on seats inside one account. It ran on borders. Target the enterprise first, Kumar said, “and once we acquire them we go to multiple countries where they are working with the farmers” — mid-market comes later.

Customer comes with 10,000. In three months they expand 200k, because they proven the product in one location, they take it to 20 location.

Krishna Kumar, founder & CEO, Cropin

Kumar put revenue churn at 1% a month, though he called it “value churn” and Nathan had to walk him through the arithmetic on air: on a $100,000 customer, $1,000 of revenue leaves. On net revenue retention he was less sure. Asked whether expansion inside a fixed cohort outran that churn, he offered the $10K-to-$200K example, then conceded: “I need to check the numbers on this one.” Nathan said he was surprised the question had not come up during a Series B closed six weeks earlier.

Why a $13,000 contract is worth chasing in person

The number that makes the whole model cohere is the one Kumar gave last. Cropin sold with a mix of inside sales and feet on the ground, aimed at enterprise accounts, at a customer acquisition cost “close to five thousand dollar to six thousand dollars” and an eight-month payback. Contracts were billed quarterly, because the buyers are enterprises.

Set that against Western CAC benchmarks and it barely reads as the same business. A field sales motion into named enterprise accounts is expensive almost everywhere; Cropin ran one against a $13,000 average domestic contract and still recovered its cost in eight months, because the salesperson doing the running was on a Bangalore cost base too. That is the real arbitrage in emerging-market vertical SaaS — not cheaper software, but a go-to-market that survives at contract sizes a US cost base could not touch. And the accounts it lands are precisely the ones that start paying international prices the moment they expand abroad.

What the database recorded next

The dataset does not record whether the $4.5 million landed by March 31 — the next dated revenue row is nearly two years later. It reads $18 million as of the end of January 2021 — seven and a half times the run rate on tape. The most recent revenue figure, dated June 2024, is $63 million, and the database flags it as an estimate rather than a founder-reported number, which is how it should be read.

What is recorded more firmly is the capital. Cropin’s funding row, dated January 2021, puts total raised at $20 million — $8 million above the $12 million Kumar cited at taping, with nothing recorded since. Against the estimated $63 million, that is roughly three dollars of revenue for every dollar raised, on a company that took nine years to reach its first $2.4 million. Efficiency of that shape is unusual, and it is available for the same reason a $5,000 acquisition cost worked against a $13,000 contract.

Headcount tells it from the other side. Kumar had 150 people at taping, 90% of them in Bangalore. The database records 260 by January 2022, then a slow decline — 251 at the end of 2022, 245 in July 2023, 224 in October 2024. If the $63 million estimate is anywhere near right, revenue per employee went from about $16,000 at taping to about $280,000 on a headcount that peaked three years ago. That is what the AI spending Kumar described in 2019 was meant to buy. The current figures sit on Cropin’s GetLatka profile.

Nathan closed with the famous five. Favourite book: “Straight from the Gut,” the memoir of GE’s Jack Welch — a GE man to the end. CEO he follows: Marc Benioff. Favourite tool: Pipedrive. Six to seven hours of sleep, no kids yet. And then, asked what he wished his 20-year-old self knew, the engineer who left GE to sell software to potato growers answered with the thing that had let him leave: “there’s a world well beyond engineering, because India mostly either they want to become an IAS officer or engineers.”

SourcesNathan Latka’s January 2019 interview with Krishna Kumar, founder and CEO of Cropin; GetLatka dataset rows dated January 2019 through October 2024, with the May 2024 revenue figure flagged in the database as an estimate.