AgTech Retro #

Aspiring agtech entrepreneurs often ask about my lessons learned from 5 years of co-founding and operating Tule. Below are some of my most common answers.

First principles #

Entrepreneurs could save themselves many years of heartache if they would think through their agtech idea from first principles up front. A few key examples:

Quantify Maximum Potential Upside #

Most technologists assume there is a lot of room to optimize efficiency in farming. This is often false; farmers have gotten pretty good at their craft over thousands of years. Entrepreneurs should quantify the maximum potential improvement their technology could manifest and then calculate the ROI on that improvement.

For example, if you are starting an irrigation optimization company, you should calculate:

How much are farmers over watering or under watering with their current method?
If they irrigated perfectly, how much would it improve yield or cost per pound?

If the maximum potential upside isn’t compelling, an imperfect solution isn’t compelling.

Does It Obey the Laws of Physics? #

Often technologists get so excited about a certain technology, that they overlook if the fundamental capabilities of those technologies are capable of solving agronomic problems. For example:

For drones and satellites, does the imagery spectrum data have the agronomic signal, in the necessary timeframe, for it to be actionable?
For indoor farming, does the energy budget for solar and light bulbs make sense compared to free sunlight?

What Is The Limiting Factor? #

Yield is usually limited by one of nutrients, water, solar energy, or pest pressure. Make sure you are solving the limiting factor.

Define Truth #

Many agtech startups claim to measure something. Soil moisture probes, weather stations, NDVI imagery, etc. But when you ask for technical specifications on accuracy gurantees and error bar details, often they can’t be provided. Usually there is a good reason for this, such as heterogeneity in the ecosystem (soil, wind, slope, etc) impacts accuracy. Sometimes more research is needed to answer these questions. Regardless, it is worth understanding the details.

The better you can define “truth”, the better you can improve your technology to match truth. Also, the easier it is to prove to customers that your technology is working and can be trusted.

Separate Economy #

Very few people appreciate just how unique and separated the agriculture industry is from every other industry. Many investors roll their eyes at this statement, as most entrepreneurs believe they are a unique snowflake. But it is actually true in agriculture, and anyone who doesn’t heed that truth will have a very tough journey. A few examples:

Different universities: Agriculture has different elite universities. In ag, a degree from UC Davis is much more respected than Stanford or Harvard. The schools that are respected in agriculture are historically land grant universities.
Land Zoning: Agriculture has its own land-use zoning.
Regulators: Things like pesticides are regulated by the Ag Commissioner. Other public offices such as City Councils are strictly not allowed to pass laws that conflict with the Ag Commissioner’s purview.
Banks: There are special banks that farmers bank with, with specialized subsidized loans. For example, Rabobank.
Sales Channels: Almost all agricultural purchases are made through either industry trade shows or farm supply distributors. The Internet playbook doesn’t work, yet.
Taxes: Farmers have a different set of tax exemptions vs other industries.
Clothes: Think boots, belts, hats, etc that are culturally unique to farmers.
Music: Country music tells the cultural story of farmers and ranchers.
Business Timelines: Farmers make incredibly long-term business decisions. Orchards take 3 years to establish before the first crop. Vineyards need to be replanted every 20-25 years. Land is passed from generation to generation.
Seasonality: The original “seasonal business”; plants take time to grow.
Rights: Agricultural water rights often supercede municipal government rights.
Religion: Many industry events begin with an opening prayer and the National Anthem.
National Security: Food security is a national concern, impacting international trade.

Because this is so different, farmers have a strong distrust of “city people”. They will say phrases like, “if you didn’t grow up with it, you can’t understand it”. So, anyone starting an agtech company should make sure they have some credible liason to bridge these two distinct worlds. Otherwise, good luck getting a meeting ;)

Heterogeneity #

Farming is not one industry. It is a bunch of small markets grouped together. People who grow vineyards vs strawberries would not consider themselves part of the same industry. They have different agronomic farming practices. They attend different industry events. They buy different specialized equipment. They have different distribution channels and economics.

Very few products are applicable in “all farming”. Beware of a TAM slide that says, “there are X million acres of farmland in the world”. Clarify the subset where the technology is applicable and the economics make sense.

Economics #

Silicon Valley people are used to “non-zero sum thinking” or “growing TAM”. Conventional wisdom is that you can always add more value and raise prices.

By contrast, an almond farmer friend described his job as “loss minimization”. Each year when his trees bloom, it sets the upper limit on his profit for that year. Each unpolinated blossom is loss. Each bud lost to frost is loss. Each nut eaten by a pest is loss. Unlike software, there are biological limits to the upside maximization of farming.

Of course farmers can increase their revenue by farming more acres. This is how tractor manufacturers justify the cost of equipment. But farmers really think about revenue and costs on a per acre basis.

If you understand the asymetric risk the farmer faces, you will understand how to serve them better. Avoiding killing the plants is much more important than achieving a small increase in yield.

Long-Term Planning #

Farmers think on much longer timelines than tech people. For example:

When planting a vineyard, farmers expect a 20 year life of the vines, and the vines don’t produce much fruit for the first 3 years.
Farmland is purchased with the expectation of holding it for multiple generations.

Market concentration #

Agriculture is a very concentrated market. In the wine grape and almond markets, most of the acreage is managed by the top 200 farmers. If you want to build a profitable business that impacts a large amount of farmland you need to focus on these large growers.

Stakeholders #

Large farming companies have multiple stakeholders that you must consider as part of the sales process. It isn’t a generalist job of a single person who works the land. For example, in vineyards there is:

land owner
vineyard management company hired to manage the land
viticulturist (expert in plant disease, nutrition, and irrigation)
laborers (who do the physical work)
scounts (who drive around checking on things daily)
irrigators (people who turn on and off the water)
etc

Where the margin is #

Farming itself is a low margin business selling a commodity product. There are higher margins in:

Inputs (fertilizer)
Processed products with brands (wines, dried nuts, etc)