The benefits – and perils – of commercializing biometric technology. Biometric technology used to be exclusive to government administration – passports that store fingerprints and faces, or criminal databases that collect the DNA information of people who have been arrested. In recent [...]
CropIn Co-founder and COO Kunal Prasad talks to Jumpstart about how the startup solves three key agricultural issues.
The agricultural sector in India accounts for nearly 42% of total employment. Yet, it only makes up about 15% of the country’s gross domestic product (GDP). Indian farmers are responsible for growing the crops consumed by over 1 billion people in India alone. However, the vast majority of them remain poor, with many depending on non-farming income sources to escape poverty.
This is because 85% of Indian farmers own farm lands of five acres or less, making the agricultural sector highly fragmented. These small farmers are susceptible to production risks including drought, floods, lack of adequate financing, large yield gaps, lack of assured and adequate irrigation, and crop failure.
Further, 30-40 % of total production is wasted at various stages of the supply chain, aggravating the farmers’ stress. So, in 2010, two childhood friends set out to tackle the challenges in the agricultural sector by putting technology to work. The result was the establishment of CropIn, an agritech startup that uses artificial intelligence, machine learning, satellite monitoring, and weather analysis to help farmers improve crop yield.
According to CropIn Founders Krishna Kumar and Kunal Prasad, the startup has reached $10 million in annual recurring revenue (ARR) and is growing 3X year-on-year. Last month, the startup raised $20 million in Series C funding led by ABC World Asia, a private equity firm focused on impact investing. CropIn also counts the Bill and Melinda Gates Foundation, Chiratae Ventures, and CDC Group among its investors.
Experimenting to find the right business model
Kumar and Prasad were deeply affected by the farmers’ grievances. Their first move was to try to solve the farmers’ problems by utilizing technology and providing advisory services.
The farmers were to pay CropIn for advisory services after a crop season. After trying this model for the first season, Kumar and Prasad realized that the farmers would not be able to pay for such services. This initial business model would not be scalable.
In 2011, CropIn pivoted to a business-to-business (B2B) model, and targeted companies, government bodies, and organizations including PepsiCo, McCain Foods, BigBasket, the state government of Karnataka, the central government of India, and the World Bank.
“We realized that [the companies] also were in [need] of a real time understanding of what’s happening on their farms, because these organizations were working with thousands of farmers on the ground, growing different kinds of crops for them, which they would then source,” says Prasad.
The companies did not have real time data about sowing, area of lands, locations of the plots, pest infections, harvest time, or overall supply, says Prasad. All this information was collected manually by extension workers deployed by the companies on the field, he adds.
“We solved that problem by really building an integrated digital platform, which could connect the farmers, the extension workers, and the businesses on a single seamless platform,” says Prasad. The platform is offered as software-as-a-service (SaaS) on a subscription basis.
Simply put, farming companies pay for CropIn’s services to increase productivity and traceability. This results in an increase in yields and efficiency that increases farmers’ revenue while giving companies greater visibility and control over their farming assets, and greater predictability for harvests.
CropIn uses technology to solve three key issues
According to Prasad, the agriculture sector has three major needs which CropIn strives to address: maximizing production to increase revenue per acre of land, cutting down the cost of production and financing, and risk management.
Increase productivity and lower production costs
“In India, the production standards are 40% to 50% of the global average – we are far behind the curve,” says Prasad. “So we use technology to pass on real time solutions and advisory to the farmers, which when they adopt, can actually increase their productivity.”
But Prasad cautions that increasing productivity is not enough to increase the revenue of farmers, as there is a need to lower production costs at the same time. CropIn does this with the help of “productive, proactive, and predictive advisories” that can substantially bring down the cost of production.
For instance, in a chilli farm, the typical practice is to spray insecticides and pesticides every three to four days, regardless of whether there is a pest problem or not, says Prasad. The cost of these chemicals increases the cost of production and lowers the quality of crops.
“We have models where we can predict certain pest and disease conditions to proactively advise the farmers,” says Prasad. CropIn’s predictive model provides insights that include the right sowing window, the irrigation schedule, fertilizer and chemical usage times, early warnings of pest infestations, and harvest period estimation – indicators that if heeded, are likely to lower production costs and increase productivity.
Decreasing costs using AI, satellite, and historical data
These predictive models are based on remote sensing data that is collected from satellites and historical climate and crop stress data. The satellite images are analyzed to understand the particular crops being grown in a particular plot, the stress level of the crop, the yield of the plot, and so on. This data is then used by machine learning algorithms and deep learning models to build predictive models that provide weather-based advisory to farmers.
The historical stress level of a particular plot is further analyzed using AI-powered predictive models. For instance, in an area prone to drought, the historical stress level of crops would be high. But by using by using AI insights to practice timely sowing and harvesting, Prasad says the stress level of the plot can be reduced or the crops salvaged.
All these models are integrated into a mobile application. The app helps with “managing farmers, managing extension workers, managing crops, digitizing farming, the land records, the crops grown,” and more, says Prasad. In other words, the app serves two purposes – collecting information to digitize farming, and providing actionable insights to farms to improve efficiency.
Moreover, the app has two beneficiaries: the farmers, and the farming companies. By providing early warnings about pest infestations and weather conditions, the app enables farmers to improve their productivity.
On the B2B side, CropIn’s app allows farming companies’ “extension workers” to manage hundreds of farmers using a mobile-based real-time application. They can use the app to collect data, disseminate information, organize safe chemical distribution, or arrange harvest collection services for farmers.
Although CropIn uses remote sensing data, it also uses these extension workers to collect primary data in the field. There are currently 30,000 extension workers using CropIn’s platform.
“These extension workers basically do day-to-day visits of the farms, they register the farmer, the crop that the farms are growing, the location, the survey, the land or the property, crop pictures,” and all the data is uploaded on the CropIn platform, says Prasad.
Besides increasing productivity, CropIn also wants Indian produce to have consistent quality and traceability in order to enter international markets. To that end, the startup has also developed a smart supply chain solution called SmartWare.
SmartWare tracks the produce as it is collected from the farmer by CropIn, stored in warehouses, sorted and graded, and tagged with a stock keeping unit (SKU) and sold with a QR code. This QR code can be scanned by any customer in any country and it shows the location where the product was grown and which farming company or farmer grew it.
Financial risk management
The cost of financing also adds to farmers’ burdens, since small farmers usually borrow from local moneylenders at exorbitant interest rates of 50-60%, says Prasad. Therefore, CropIn’s risk management model, SmartRisk, offers banks information about the farmers’ landholdings, their creditworthiness, and a plot level monitoring system.
Since many farmers don’t have a paper trail to establish creditworthiness the traditional way, SmartRisk evaluates them using the historical crop data of the plot and the estimated yield and revenue of the farmer, reducing the risk of lending for financial institutions. This allows farmers to obtain financing at a 6-12% interest rate, significantly bringing down financing costs.
For its corporate clients, CropIn uses AI and ML to forecast risks such as crop failure, low yield, or low-quality harvest. These overall forecasts can help clients reassess their strategies for upcoming quarters or seasons. For instance, if the harvest prediction for a particular crop is low in a particular region, companies can shift their focus elsewhere and improve their return on investment.
CropIn plans to expand aggressively this year
So far, CropIn says it has worked with over 4 million farmers through its platform. 70% of the farmers are located in India, most of whom are smallholders. The rest of the farmers are spread across more than 52 countries in Asia, Africa, Latin America, parts of Europe, and the U.S. With the fresh funding raised last month, Prasad says the startup plans to expand its geographical presence globally.
Moreover, CropIn supports over 9,000 different crop varieties, and has a diverse customer base including farming contract companies, seed companies, tractor companies, government bodies, and developmental organizations, among others.
The startup’s growth so far is testament to the need for technological solutions in agriculture, and its technology is an indication of what frontier technologies like artificial intelligence can do when put to work solving problems at the grassroots.
Header image courtesy of wilsan u on Unsplash.