The numbers Brief Episode map Timeline Takeaways Charts Quotes Clips Glossary Go next Transcript

Episode 67 · Impact · 92 min

Horticulture as the second software export

India's software exports run to about $113 billion; its horticulture exports are nowhere close, despite 30 million hectares that already yield some 300 million tonnes a year. Fasal's claim is that the gap is not soil or climate but decision-making: the four or five calls a farmer makes every day on water, spray and fertiliser are still governed by inherited wisdom the weather no longer honours. Its answer is a ₹14,000 sensor priced like a Maruti 800, already live on 50,000 acres.

Supply chain & agri Regulation & policy Deep tech & hardware

Shailendra Tiwari

Co-founder, Fasal · with Vishal Krishna

1:31:46

Said in this episode

▶ 4:09

$113 bn

India's software exports, the benchmark

Tiwari's yardstick for what an optimised industry looks like; he argues horticulture exports are nowhere close and could be the second one. Vishal cites a higher $150 billion figure in the same exchange.

▶ 12:34

30M hectares

Land under horticulture in India

The long-run addressable base, against roughly 300 million tonnes of horticultural output a year, as stated in the opening.

▶ 11:14

90 of 100 L

Groundwater that goes to agriculture

A 2020 World Bank figure as cited on air; Tiwari's counter-estimate is that crops need something closer to 30–50 litres, making India's water shortage partly self-inflicted.

▶ 13:04

50,000 acres

Farmland running on Fasal today

Built over about three years since 2018, with a stated target of half a million acres in the next 12–15 months and 30 million hectares over the long run.

▶ 47:50

₹14–30k

Price of a unit, per 3–4 acres

Three-to-five-year hardware life, under 1% of a farmer's annual spend; farmers pay 40% in advance and EMI was being launched at the time of recording.

▶ 45:57

40–60 districts

Where 80% of horticulture land sits

Fasal's own clustering estimate, and the reason its expansion is district-by-district: about 15 districts opened last year, roughly 45 this year.

The brief

The argument in sixty seconds

Tiwari's claim is that Indian horticulture is a software industry nobody has written yet: India's software exports are about $113 billion and its fruit-and-vegetable exports are nowhere near, despite 30 million hectares already producing roughly 300 million tonnes a year. The bottleneck, he argues, is not agro-climate (India has the zones, the soil, the resources) but the four or five decisions a farmer takes daily on irrigation, disease and pest, fertiliser and now climate risk, decisions still made on wisdom inherited from generations that the new weather has stopped validating. When the wisdom fails the decisions fail, when decisions fail the crop fails, and the livelihood goes with it. Fasal's answer is deliberately unfashionable: an IoT device standing in the field rather than drone or satellite imagery, because a farmer needs prediction rather than detection and you cannot fly a drone continuously. It is priced at Maruti 800 levels, ₹14,000 to ₹30,000 for a unit covering three to four acres, under 1% of what a farmer spends in a year, because 80% of the world's farmers are smallholders and anything dearer is a demo. After Galwan the team refused Chinese components and built the hardware in India; 50,000 acres are live, farmers pay 40% up front, and payback lands inside a season. The wider argument is political: 90 of every 100 litres withdrawn from the ground goes to agriculture, ₹1 lakh crore a year goes to fertiliser subsidy, and the trading layer everyone calls villainous is mostly arbitraging information that WhatsApp has already destroyed. Get the daily decisions right, Tiwari says, and the water table, the subsidy bill and the export ledger all move together.

Worth your time if you are

Agri-tech founders choosing between imagery and ground sensors

Hardware teams pricing for the bottom 80% of a market

Policy readers arguing about fertiliser and water subsidy

Agricultural-sciences students who took the degree by default

Exporters who need predictable, traceable horticulture supply

Episode map

Where the conversation travels

Every block is a chapter, coloured by what it's about. Click any of it to jump straight to that minute on YouTube.

Supply chain & agriRegulation & policyDeep tech & hardwareUnit economicsData & digitisationLeadership & org

01Cold open: the case for farm data 0:00 Vishal sets the frame from his own family's lost farmland. India is a net agricultural exporter that still imports $10 billion of edible oil a year, produces about 300 million tonnes of horticulture, and runs on land holdings whose productivity trails China's. 02Horticulture as India's second software 2:16 Tiwari's opening thesis: India's software exports are around $113 billion and its horticulture exports are not close, so the country's agro-climatic advantage is an unexploited export industry, and optimising it would change what villages look like. 03Four decisions, made on failing wisdom 6:58 Irrigation, disease and pest, fertiliser and climate risk are the four or five calls a farmer makes daily, and Fasal exists to move them from inherited wisdom to logic, a shift that has cut spray counts 20–40% across 50,000 acres since 2018. 04Maruti 800 price, Ferrari inside 14:22 Because 80% of the world's farmers are smallholders and 'smallholder' means nothing without the crop (a paddy acre grosses ₹50,000–60,000 while a pomegranate acre can gross ₹10 lakh), Fasal capped the product at under 1% of a farmer's annual spend and made the engineering fit the price. 05Why the trader and the exporter benefit 19:37 The same farm-level data gives corporates visibility over 10,000 acres from a Mumbai desk, tells input companies where to stock before the problem arrives, and turns a four-day shelf life into seven, so value percolates through the chain rather than accruing to the app. 06What isn't on Google about Indian farming 23:37 Building the full stack meant rejecting the frail-farmer image search returns, and hiring 150 people who were told plainly that anyone joining for a higher salary was in the wrong company. 07Galwan, and why it had to be hardware 31:17 Devices that survive Rajasthan heat and Himachal snow, a conscious post-Galwan decision to source nothing from China, and the core architectural bet: continuous IoT streams over drone imagery, because agriculture needs prediction rather than detection. 08Purpose, happiness and the Dandi march 39:35 Tiwari argues purpose and happiness are inclusive rather than exclusive (a Great Place to Work certificate mattered less than a customer-farmer taking an investor call at ten at night), and that a team given direction marches to Dandi without being managed. 09Districts, ₹14,000 units and 40% advance 44:28 Horticulture is clustered: roughly 80% of the land sits in 40 to 60 districts. So Fasal opens districts crop by crop, sells self-service units at ₹14,000–30,000 with a three-to-five-year life, and finds farmers paying 40% in advance because payback lands inside a season. 10Agricultural universities and the missing layer 51:55 Too many students still take agricultural science because they missed medicine, yet Indian universities hold deep plant physiology, pathology and entomology. The missing piece was never the science but a digital delivery platform to model it and scale it. 11An MoU with IMD for nowcasting 57:20 Fasal is building region-specific nowcasting models with the India Meteorological Department: six-to-twelve-hour forecasts that respect the topology of a Shimla slope or a north Karnataka plain rather than one national model. 12Migration, narrative and the humble farmer 1:01:57 Both host and guest push back on the political-economy script of the pitiable farmer, arguing the family side of Indian farming was always celebratory and that engineers should take their data skills back to the land. 13Food security is national security 1:05:35 Why the state must intervene at all. A 90%-subsidised DAP bag, the Agriculture Infrastructure Fund paying for on-farm solar cold storage, 'more crop per drop' drip subsidies and contract farming are all read as competitiveness measures, not charity. 14Traders are not demons 1:13:00 A three-acre grape farmer spending ₹7.5 lakh to earn ₹10–20 lakh is not powerless, and the information asymmetry traders once exploited is dying in WhatsApp farmer groups, leaving value creation, not arbitrage, as the game. 151991 again: unblock, then get out 1:19:06 Tiwari reads this decade as a second 1991: the government's job is to unblock aspiration and build atmosphere, as the ₹10,000 crore startup fund of funds did, while the withdrawn farm laws were a step whose dialogue failed rather than a step in the wrong direction. 16Great-grandfather's answer, and Dickens 1:27:00 Asked what inspires him, Tiwari recounts a WWII veteran great-grandfather's instruction to raise children useful to society, and Vishal closes on the Tale of Two Cities passage the guest quoted earlier.

Timeline

The arc, briefly

Four years of building a sensor for smallholders, and the case for horticulture as an export industry.

Pre-1991India built its own hardware. Tiwari's answer to thirty years of being told the country cannot do it. He points at Siachen and at the road from Electronic City to the airport as evidence that somebody just has to start.

2018Fasal starts, and runs pilots. The premise is that a farmer makes four or five decisions a day on irrigation, disease, pest, fertiliser and now climate risk, all of them on wisdom handed down through generations that the new weather no longer validates. Building an IoT device for farmers was hard work at a time when everyone assumed farmers had no money and no interest in technology.

2018-21Three years to get the device right. Drones and imagery were explored and dropped: you cannot fly a drone continuously, and an image detects a problem that has already happened, where agriculture needs prediction. He puts it as yoga rather than the doctor. After Galwan the team decided to buy nothing from China and build the hardware here, which was the harder road.

2020The World Bank puts 90 of every 100 litres on farms. Tiwari reckons crops need 30, 40 or 50, so a good part of the water shortage is self-inflicted. The same over-application shows up in spraying done out of fear, which is why Indian horticulture struggles to reach European shelves.

2020-21Twenty people become a hundred through the COVID waves. Mission-oriented hires brought in more of the same, and units went out by courier with a video call so farmers could deploy them without anyone visiting. The company is now 150 people, and both founders still interview every one of them.

Today50,000 acres, and an MoU with the IMD. Farmers on the system cut spray counts by 20 to 40%, and cost of cultivation by as much as 60% in places. Units run ₹14,000 to ₹30,000 for three or four acres, last three to five years and come in under 1% of a farmer's annual spend, which is why 40% is paid in advance. The met department work is region-specific nowcasting, six to twelve hours ahead, built for a Shimla slope rather than one national model. Fifteen districts opened last year and about 45 this year, against a target of half a million acres in 12 to 15 months.

Takeaways

Ideas to carry out of this hour

Horticulture is an export industry India hasn't written yet

Tiwari benchmarks against the one Indian industry that scaled globally: software exports of roughly $113 billion, against horticulture exports he says are not even close. The inputs are already there: agro-climatic zones, soil, 30 million hectares under horticulture, some 300 million tonnes a year. So the deficit is optimisation, the thing software veterans did to their own industry. The pitch is not a farm app; it is that the same discipline applied to fruit and vegetables makes India the food bowl of the world.

▶ 4:09 ▶ 5:12 ▶ 5:25

The unit of change is four or five daily decisions

Map a farmer's day and it reduces to irrigation, disease, pest, fertilisation and, newly, climate risk. Those calls have always been made on wisdom handed down over generations, and Tiwari's claim is that the wisdom is now failing because the climate it encoded no longer exists: Bangalore itself is not the city he arrived in. When wisdom fails the decisions fail, the crop fails and the livelihood is at risk, so the intervention has to sit at the decision, not at the market or the subsidy.

▶ 7:28 ▶ 8:59 ▶ 9:48

Prediction beats detection, which forced the hardware

Fasal explored drones and imagery first and rejected them: you cannot fly a drone continuously, and image-based systems detect a problem that has already happened. Agriculture, Tiwari argues, needs prediction (the yoga argument rather than the doctor argument), and prediction needs a continuous stream of ground truth, which is why a software company ended up building a patented in-field device. For a one-acre farmer, a spray avoided is ₹5,000 saved, which is the whole ROI conversation.

▶ 35:40 ▶ 36:10 ▶ 36:38

Price is the specification, not the outcome

Eighty per cent of the world's farmers are smallholders, so a product that doesn't work for them is a product with a long way to go. Fasal fixed the price first (under 1% of what a farmer spends in a year, a unit at ₹14,000–30,000 covering three to four acres with a three-to-five-year life) and made it an engineering problem to deliver everything else inside that envelope. Tiwari's phrase for it: Maruti 800 pricing with what you'd expect sitting in a Ferrari. He also notes the hardware cost was said to be 400 times higher when they started, a figure worth treating as directional.

▶ 16:11 ▶ 19:21 ▶ 47:50

The water problem is an artificial shortage

Citing a 2020 World Bank figure, Tiwari says 90 of every 100 litres withdrawn from the ground in India goes to agriculture, and that crops probably need 30, 40 or 50, not 90. Add to that the over-spraying done out of fear, which is precisely what keeps Indian horticulture off European shelves, and the case is that India's water shortage and export failure share one cause. His policy extrapolation: the fertiliser subsidy runs about ₹1 lakh crore, and it shrinks if every farm doses to what the plant actually needs.

▶ 11:14 ▶ 11:30 ▶ 1:12:28

Geography does the go-to-market

Thirty million hectares sounds unservable until you notice horticulture is clustered, not spread. Fasal's own calculation puts about 80% of that land in 40 to 60 districts. So the company goes crop-by-district: Nashik for grape, Burhanpur for banana, Dhar and Barwani for chilli and tomato, Kolar for tomato. Fifteen districts opened last year, about 45 this year, with 60 as the ceiling. Marketing spend stays low because one farmer's demo sells the next. Tiwari's line is that doing good for farmers shows up as karma and EBITDA.

▶ 45:26 ▶ 46:33 ▶ 48:51

Hire for mission, and refuse the rest

The founders still interview every one of the 150 hires, and tell candidates outright that anyone joining for a higher salary is in the wrong company, and they are comfortable not hiring. Mission-driven people hired more of the same and the company went from 20 to 100 through the COVID waves. The corollary Tiwari draws is that once purpose does the motivating, attrition and inspiration stop being management problems and leadership reduces to direction: people will march to Dandi and talk about how good the roadside sleep was.

▶ 30:18 ▶ 33:02 ▶ 43:14

Traders are not the villains; asymmetry was

Both host and guest refuse the standard framing: everyone in the value chain adds something, and a three-acre grape farmer spending ₹7.5 lakh to earn ₹10–20 lakh is not a powerless party. What traders did exploit was information asymmetry, and Tiwari says one of Fasal's biggest revelations was the sheer number of farmer WhatsApp groups. A grower in Arunachal now knows the Delhi apple price. With arbitrage dying, the only lever left is value creation: what lets a farmer command ₹90 a kilo for grape instead of ₹35.

▶ 1:14:15 ▶ 1:16:14 ▶ 1:17:17

The numbers, drawn

What the episode measures

Every figure below was said on air, with timestamps included and caveats kept.

Where every 100 litres of groundwater goes

litres per 100 withdrawn

As stated on air, citing a 2020 World Bank report for the 90-litre share; the 30–50 litre range is Tiwari's own estimate of actual crop requirement, not a published figure.▶ 11:14

Why 'smallholder' means nothing without the crop

₹ lakh per acre

Upper bounds of the ranges Tiwari gives on air: paddy at ₹50,000–60,000 revenue on ₹10,000–20,000 of inputs per acre, pomegranate at up to ₹10 lakh on ₹2.5–3 lakh. His argument is that marginal, small and medium are crop- and region-contextual terms.▶ 17:03

Four calls a day, and what breaks when they miss

the decisions Fasal targets, and the cascade Tiwari describes

Tiwari's own reduction of a farmer's day, and the chain he draws from it. Fasal's claim is that the intervention has to sit at the decision rather than at the market or the subsidy.▶ 7:28

Why it had to be a box in the field

imagery against ground sensors, as Fasal weighed them

The architectural bet, in Tiwari's framing: agriculture wants the yoga argument rather than the doctor argument. The patented in-field unit is what a software company ended up shipping.▶ 35:40

Worth keeping

Lines that stay

India's software export was about $113 billion. We are excited about this space because horticulture can be India's second software — that's the potential here.

▶ 4:09

Farmers are not charity. Agriculture is not charity, where you need policies to percolate money into their pockets. You have to solve problems the way any industry has optimised itself.

▶ 29:16

It is a Maruti 800 priced product which has everything that you will feel sitting in a Ferrari.

▶ 19:21

If you are joining this company because you want a higher salary, you are landing up in the wrong place. We are very okay making that call — saying we can't hire.

▶ 30:32

Marketing costs are significantly lower, because when you do something good for farmers it reflects in the form of karma and EBITDA.

▶ 48:51

Clips that travel

Short on time? Start here

Agri-tech founders and agronomy students

The four decisions that decide a crop

The whole thesis in one stretch: irrigation, pest, fertiliser and climate risk, why inherited wisdom is failing, and the 50,000-acre evidence.

7:28 → 14:22 · 7 min ▶ Watch clip

Hardware teams pricing for the bottom 80%

Maruti 800 price, Ferrari inside

Paddy versus pomegranate economics, why 'smallholder' is a contextual term, and how a price ceiling became the product spec.

14:22 → 19:37 · 5 min ▶ Watch clip

Deep-tech founders weighing sensors against imagery

Why not drones, and why not China

The post-Galwan sourcing decision, hardware that survives Rajasthan and Himachal, and the prediction-versus-detection argument that forced IoT.

31:17 → 39:35 · 8 min ▶ Watch clip

Operators building rural go-to-market

Districts, pricing and the 40% advance

Horticulture's district clustering, self-service deployment during COVID, ₹14,000–30,000 units, and farmers paying up front.

44:28 → 50:55 · 6 min ▶ Watch clip

Anyone who blames the mandi for farm distress

Traders are not demons

Grape economics at ₹35 versus ₹90 a kilo, WhatsApp killing information asymmetry, and the shift from arbitrage to value creation.

1:13:00 → 1:19:06 · 6 min ▶ Watch clip

Glossary

The jargon, unpacked

Horticulture

The cultivation of fruits, vegetables, flowers and spices as distinct from field crops like paddy or wheat. About 30 million hectares in India, and the segment Fasal exclusively serves.

Smallholder farmer

A farmer working a small plot, roughly 80% of the world's farmers, though Tiwari argues the label is meaningless without the crop, since a pomegranate acre can gross twenty times a paddy acre.

Nowcasting

Very-short-range weather prediction, typically the next six to twelve hours. These are the region-specific models Fasal is building with IMD, meant to beat global models that ignore local topology.

IMD

The India Meteorological Department, the government's weather forecasting arm, with which Fasal signed an MoU for regional nowcasting models.

DAP

Di-ammonium phosphate, the fertiliser sold to Indian farmers at roughly ₹3,000 a bag against a heavily subsidised cost, cited as an example of intervention justified by food security.

Agriculture Infrastructure Fund

A government scheme financing infrastructure at the farm gate, such as solar-powered cold storage that lets a strawberry grower wait out a bad price day instead of dumping the harvest.

Contract farming

An arrangement where a buyer commits to a price before the season, giving the farmer certainty on realisation and the buyer visibility over aggregated, monitored supply.

Drip irrigation

Tubing that delivers water drop by drop at the plant's root rather than flooding the field. It is the technology behind the government's 'more crop per drop' subsidy programme.

Connections

If this resonated, go here next

EP 63 · Impact

Full transcript

The whole conversation, searchable

Auto-generated captions, lightly cleaned. Click a timestamp to open that moment on YouTube.