Here are five ways to make data, and information extracted from data, more valuable for farmers.
1. Define a clear unique selling point (USP), making a conscous choice between providing a platform nd data/intelligence.
The digital ag tech space is diverse with startups addressing several different aspects including farm management systems (FMS), camera/sensor hardware, data intelligence, and drones and robotics. But in general, ag tech companies are not well-differentiated, as evidenced by drone manufacturers trying to develop data analytics services, FMS providers trying to add a data intelligence layer, data analytics and precision farming companies trying to push their own version of an FMS, agri machinery producers trying to lock customers into their own solutions, etc. Similar data is available to multiple players, and the majority of ag tech start-ups use off-the-shelf technologies, such as multispectral imaging and high-resolution RGB cameras. Furthermore, large agricultural corporations, such as Monsanto, Bayer, and Syngenta, have proprietary data sets. The industry of digital technologies in agriculture is not yet consolidated and there are no clear market leaders.
Especially in this context, we believe that it’s extremely important to define the USP early; make a clear differentiation of your products and services, and make a conscious choice whether to be primarily a platform or a data intelligence provider. The distinction is not definitive, but focus is necessary. Platform creators should center on the seamless integration of multiple data sources, connectivity of the data with the farming equipment of various providers, visualization, and usability of the FMS. Data intelligence providers should work to build unique data sets and algorithms that can be used by multiple platform providers. When a data intelligence provider needs to create a platform to deliver and visualize results and to integrate data sources in order to create the intelligence product, these activities should support the main goal of providing intelligence to customers.
At Gamaya, we have positioned ourselves as a crop intelligence provider based on our strengths such as our hyperspectral imaging system, agronomy insights, and machine learning. We provide the information about where a problem is, and also offer diagnosis of the issue and collaborate with the farm manager and agronomists to provide a potential treatment plan. Being a data intelligence provider, we are able to plug our data into other FMS or platforms using an API-based interface.
2. Deeply integrate agronomy with digital technologies to interpret the data from a farmers perspective.
The majority of ag tech startups seem to see the ag tech industry from the perspective of the tech industry. However, in ag tech, ag is the main component, meaning that ag tech is driven by the nature of the agricultural industry. The nature of the agricultural industry depends on nature itself: weather, soil type, precipitation rates, disease burden levels and other external, and often uncontrollable, factors influence a farmer’s end result. This necessitates the use of very local and not easily scalable agronomy practices and crop- and region-specific operations. From a business perspective, the agricultural industry has a long sales cycle; fragmentation of some markets; dependence on personal relationships between agronomists and farmers; highly volatile market prices, etc.
The most successful ag tech startups are the ones that consider agronomy and agricultural expertise as a key element in building a product that will be demanded by farmers. Integrating agronomy knowledge, commercial agriculture best practices, and farming know-how into the product development process can help to narrow the gap between digital technologies and agriculture.
Connecting agronomy expertise with technology can facilitate the process of interpreting the data. Very often data interpretation is left to farmers. In these cases, farmers equipped with high-resolution Red-Green-Blue (RGB) spectrum cameras or Normalized Difference Vegetation Index (NDVI) maps illustrating plant health from drone or satellite imagery, and other derivative products from multispectral imaging, can be left struggling to find the agricultural importance and insights in the technical data. Though very strong in agronomy and agricultural operations, often a grower’s background is not in image and data processing and interpretation. Many providers of these RGB and NDVI products don’t provide the interpretation because their expertise is not in agriculture. Building a bridge between the Ag and the Tech provides the most value: farmers don’t need data—they want solutions to their farming problems.
3. Make a distinction between data, information and knowledge
Data, information and knowledge are three completely separate categories that can be mathematically defined according to information theory. The difference between data and information is that information is created or extracted from data by putting the data into a context; in the case of agriculture, the context is environmental/agricultural. The difference between information and knowledge is the existence of a purpose defined by society. It’s impossible to define knowledge without a purpose. Applied to agriculture, knowledge is actionable information that can be used to make certain treatment actions, such as spraying chemicals, spreading fertilizers, optimizing harvesting time etc. Information that cannot be used by farmers to implement a treatment action—that can’t be utilized in practice—is not considered to be knowledge. As per the second recommendation above, the need to connect agronomy with digital technologies puts data into the context of agriculture and makes the extracted information valuable and actionable for farmers, thus making it knowledge.
Technology is a toolkit that allows one to speed up the process of converting data into information. Artificial intelligence is a great example of such technology. However artificial intelligence can’t convert information into knowledge without human intelligence. Going forward, the biggest innovations will come from the ability to convert information into knowledge.
Hyperspectral imaging is one of the most detailed and expansive sources of remote sensing data. We’ve developed our own patented hyperspectral camera and this has provided us early access to high-quality data to develop crop intelligence by converting information into knowledge. We use AI and machine learning techniques on these hyperspectral and other agricultural data types to create information for the purpose of improving the outcomes of agricultural endeavors.
4. Customize products to the specific needs of a particular market segment, crop and region.
The ag in ag tech typically makes products less universal and scalable since agriculture is very diverse, local, and crop- and region-specific. NDVI is one of the few universal products in agriculture, however it provides quite basic information about the general condition of the crop and serves mainly as a tool for farmers to determine where to scout for problems. When it comes to aiding the diagnosis of particular crop issues, then a product cannot be a universal broad brush because the causes of the issue likely involve multiple factors, such as crop type, variety or cultivar, soil type, region, weather, etc.
Because customization is almost inevitable for high-value products, it makes financial sense to customize a product for a particular market segment that is characterized by similar growing practices, region, climate and soil conditions. Such products provide growers in the region with results specific to their situation. Identifying large regions of homogeneous agricultural characteristics allows for the scalability needed for ag tech businesses to be viable. Therefore, customizing at the regional level balances the agricultural need for customization with the business need for scalability.
When customizing a product, ag tech companies would be wise to train their models for each crop, region, and market segment. We focus on a few mid and high-value crops and high-impact issues — diseases, weeds, nutrient deficiencies — in consolidated regions, studying growers’ problems in detail with them. Doing so usually takes at least one growing season to build the first version of the working model and sometimes requires additional seasons to update it. Agriculture, after all, is ever changing and so should be the models that describe it.
5. Focus initially on the technology-savvy and innovative growers to bring the product the market.
Customization (recommendation 4) is labor and capital-intensive and therefore must be done strategically. Part of this strategy includes identifying which specific growers to customize the product to: growers who can participate in the development and are representative of growers in the region.
Due to the need to develop a product customized to a specific crop and region, initially it is better to address the most innovative and technologically savvy growers. In general, this is about 10% of the growers. These growers can significantly facilitate the product development and data interpretation processes by providing specific agricultural knowledge, making connections between the information and possible treatment actions and by designing field trials.
Where are these innovative, tech savvy growers? There are different types of growers and agricultural players in each region. Generally, we define the three following customer segments of growers of commodity crops:
- Very large industrial growers with strong in-house capabilities: they have their own advanced technology and agronomy know-how and sophisticated agricultural machinery.
- Large growers, typically growers with more than 20,000 hectares, who have their own technology and agronomy capabilities, but also rely on external agronomy and technology support.
- Medium and small growers that have very limited internal agronomy and technological capabilities. They are often part of cooperatives that consolidate their resources.
Large industrial growers have the financial means to support product development, have dedicated agronomy and technology/ innovation departments, and are operated by business-oriented top-managers. They tend to be local champions and early adopters. Furthermore, their typical size indicates that they operate on a large percent of the farmland in the target region, making them, therefore, representative.
Starting with the most innovative and technology advanced growers, whose lands and crops are representative of the local region, creates a market share and provides a reference to market the product to growers in other customer segments. Word of mouth referrals are very powerful in the agricultural industry.
Our go-to-market strategy is designed to address each customer segment differently and we target partners and clients who are local champions and innovators. Their agronomy expertise facilitates interpretation of the data and, paired with their high technical sophistication, shapes our product development efforts. And they can use their high-tech agricultural machinery to implement treatment actions — variable rate spraying, spreading, replanting, yield maps, etc. — created by our products.
We believe that there is a tremendous untapped potential in ag tech products and services when they are properly applied to serve farmers’ needs and customized to address specific stresses in specific crops in each region.