Discover more from Agribusiness Matters
Saturday Sprouting Reads (FullStacker's Dilemma, Mineral.ai, Irrigation Tech, Covantis, Food Systems)
Greetings from Hyderabad, India! Welcome to Saturday Sprouting Reads!
Systems Thinking in Agriculture in an age of runaway Climate Change
About Sprouting Reads
If you've ever grown food in your kitchen garden like me, you would realize the importance of letting seeds germinate sooner than later. As much as I would like to include sprouting as an essential process for the raw foods that my body loves to experiment with, I am keen to see how this mindful practice could be adapted to the food that my mind consumes.
You see, comprehension is as much biological as digestion is.
And so, once in a while, I want to look at a bunch of articles or reports closely and chew over them. I may or may not have a long-form narrative take on it, but I want to meditate slowly on them so that those among you who are deeply thinking about agriculture could ruminate on them as slowly as wise cows do. Who knows? Perhaps, you may end up seeing them differently.
During one of my agritech samvaad (dialogue) events with my dear mentor friend Jagadeesh, Jagadeesh made an important point:
As Jagadeesh put it,
“Every solution we are giving to the farmer is very short in duration for that crop. We like tell him how much more yield he will get at the end of the crop and solve his problem. If you look from farmer’s perspective, he is not looking at one crop. He is looking at how finally his life is going to be at the end of three years. Will I get a higher price is all he is asking.”
Here is the fundamental curse of being a Fullstacker in Agritech: You are pouring jet fuel of venture capital into a system in which the weakest link, the farmer, is looking, in all probability, at a longer time horizon than your patient capital.
There is a deeper reason if you probe into a farmer’s relationship with time. If you ask a farmer why is he continuing to sow when the prices fell over a particular year, he would most likely have a theory that over three years, the prices would come back to equilibrium and he is, therefore, willing to take the risk.
The point here is not to defend the status quo. Yes, we need disruptive innovations to try out all kinds of marketplace experiments. But what about the time horizon?
Can you play really long-term games when you are run by jet fuel?
Agriculture, from my limited observation, is a peculiar domain where Nature respects Time more than money. If you invest money into a factory, the result is far more certain than if you invest money into a farm, because, the infrastructure in the former is created to respect money more than the latter.
This might sound strangely ironic in a blog that is titled, Agribusiness Matters. Agriculture and Agribusiness are the only domain (as far as I know) where the oldest economic adage, “Time = Money”, stands on shaky grounds.
When you view Mineral.ai through the lens of Palantir, Mineral’s unspelt product commandments become loud and clear. (Bonus points if you are able to read this in the voice of Moses)
→ Thou shall stay away from last-mile problems. Let them be the concern of our partners who will work on the ground with farmers and retailers and other players in the last mile of the value chain.
→ Thou shall not build separate products for different players in the industry. In working with the partners, we will learn how to build models that let us absorb the complexity of the plant world and convert them into repeatable technologies.
→ In learning how to build models for the complexity of agriculture, thou shall model the distributions of the data and model the dependencies in those data, mimic them as much as possible into a model that we can sample afterwards, from which we can derive newer data sets. (Check this for a detailed illustration of this point)
When you understand this densely packed product philosophy, it naturally springs forth an important question:
How do we understand Mineral.ai from an agricultural lens? What is it about agriculture that necessitates this product philosophy to solve computational agriculture at scale?
At the end of the month of February, I did a panel discussion with diverse members of the agriculture and development ecosystem. The video has come out well and I wanted to highlight this excerpt in which I deliberately cornered a panellist and an interesting exchange ensued.
Venky: "If subsidies can kill technological efficiency and create so much havoc in the behaviours of farmers, why do companies still prefer philanthropy to solve a lot of social problems?”
Rathish: It's a tricky question. I am going to answer in two different parts. There is a fundamental first-principles question. How do we treat a farmer? Are we treating a farmer as a free business agent - like any person in the market who transacts with companies?
Or are we treating a farmer as a national asset - They are working on public good, on a business that is inherently unsustainable for small farmers. But they have to do what they have to do because we have to eat.
Venky: And we've glamorized them as "annadatas"[as Indian farmers are revered as divine food producers].
Rathish: "I think we are stuck somewhere in between. Any conversation about them is largely emotional than financial. And we've put all the constraints around them. It's regulated. No export.
Now let me come to the question of philanthropy....There are places where philanthropy has no role to play....subsidies have no role to play. Basically, philanthropy addresses market failure or government failure. Where the market can't deliver, where the government can't deliver, philanthropy has to come.
Drip Irrigation is a classic example where the market could have played and solved the problem. Philanthropy should not have a role to play there.
Today, honestly, subsidies for fertilizers should not be given. There was a window of time when subsidies in fertilizers were needed to create a habit. Now it has become an addiction. Hence It is necessary to remember..where subsidies are relevant..where market failures happen.."
This is a welcome initiative as I’ve documented before, agritech tends to subsume irrigation tech under precision agriculture and since water is not priced, it has become a difficult selling proposition to convince farmers of the potential of irrigation tech. This largely reflects in the negligible venture capital activity in irrigation tech.
The report starts with an important question - Why aren’t there many partnership announcements between industry incumbents and startups?
“There is currently a disconnect between industry incumbents and startups that may be explained by differences in their basic business models.
Irrigation equipment manufacturers maintain extensive distribution networks to allow them to sell hardware and supplies such as irrigation machinery, pumps, systems, and other related accessories. For farmers, these are infrequent, larger purchases with very little after-sales service or support.
Currently, irrigation software solutions bundled with hardware are not an important revenue stream for any of the industry incumbents. Irrigation startups are developing precision ag solutions such as irrigation scheduling applications and IoT sensors to support automation and decision-making. The reach of these companies is much less than that of incumbents.
They often sell direct to consumers. For many companies, the goal of sales is recurring revenue through software subscriptions or service provision, with a focus on upselling.”
They go on to identify three trends that are driving irrigation tech:
1→ Revenue strategies have transitioned from irrigation hardware to software to services.
2→ Bundling and vertical integration of products/services to address multiple customer pain points is increasingly common.
3→ Non-governmental organizations and aid agencies are subsidizing customer acquisition costs
The report doesn’t delve into precision agriculture gameplays in irrigation tech and the subsidy incentives which end up distorting the market price signals. I wish it had also delved into water user associations and the irrigation tech happening around them. That said, this is a good start to delve into an important subsector of agritech that often gets less coverage.
I haven’t been actively tracking the agritech gameplays of large agricultural trading players. And so you could imagine my excitement when I discovered Covantis platform that was launched in February 2021 - An Avengers Style Unified Platform Entity that brings together six of the world’s leading agricultural trading companies (Marubeni later joined them in 2022) to solve the post-trade execution challenges through Blockchain.
The introductory video makes it clear which processes the platform would cover under its scope.
In this interview with CEO of Covantis, Petya shares an interesting dilemma faced by anyone who ventures out to build an industry-wide platform: Do we build the platform or do we build the network?
As Petya elaborates,
“Our value depends on our network. We lose value if participants are absent from our platform. We have spent a significant amount of our time and energy building the network. A trading house cannot use the platform to execute a trade if its counterparty is not on it. Successful networks are considered those that have at least 80 per cent of the market participants using the platform.
It will be challenging to get to 100 per cent. Some people will always want to continue doing things as they have done in the past, but we have successfully onboarded most of the trading companies in the markets where we are live.”
It’s interesting to see how Covantis had to build capabilities that allowed traders to send structured instructions to non-platform participants. Covantis has grown manifold - ‘number of legal entities and teams active in the Covantis platform grew by 73 per cent in 2022’ and ‘executed 519 million mt on the platform in 2022, its second year of operation – an increase of 246 per cent versus 2021.
I plan to do a deep dive on Covantis.io and explore its agritech gameplay in detail.
“One of the main reasons why the competition and anti-trust authorities gave us approval was that no one has access to data other than their own. There is no sharing of information stored on Covantis, nor can Covantis JV be used as a forum for the prohibited information exchange between shareholders.” - Petya, CEO, Covantis.io
That our global food system is broken and stressed is known. How do we communicate the underlying structural logic of our food systems in a way that inspires change to redesign and re-organize the system as a whole?
A new paper from Si Global Club audaciously attempts to publish a manifesto that articulates the need for food systems change in a simple, but powerful manner.
Although the paper makes you wonder if there is a certain amount of nostalgia for small-is-beautiful-localism arguments, the paper distils a few interesting nuggets to describe why our food system is broken the way it is. It would have been nicer to present them in a non-manifesto tone and corroborate a few underlying system architecture elements with data points.
I really found these points excellent conversation starters on the design of food systems.
→ “. Environmentally, an input-output logic [emphasis mine] to farming results in a depletion of the overall ecosystem's capacity to maintain itself and continue providing the ecosystem services required, e.g. soil erosion”’
“We believe what we should be aiming for is, a system where firstly people know their nutritional needs to maintain good health and are incentivized to meet them, themselves; we are not providers, we are simply enablers of this. “
“Key to the food system transition is moving away from the economies of scale of the industrial age - that worked to concentrate and centralize - to economies of organization via information networks that work to re-empower end users. This is about our newfound capacity to aggregate many small contributions rather than always being dependent upon large centralized production and distribution. In the context of food production, this starts with looking more at decentralized micro-production. This requires a food platform for the aggregation & exchange of food between producers & consumers.”
The essence of the food system transition is much more complex than moving from economics of scale to economics of variety.
Is it possible to move away from economies of scale design patterns inherent in our food system design? It’s a fascinating question that is worth exploring.
So, what do you think?
How happy are you with today’s edition? I would love to get your candid feedback. Your feedback will be anonymous. Two questions. 1 Minute. Thanks.🙏
💗 If you like “Agribusiness Matters”, please click on Like at the bottom and share it with your friend.