Insurtech | Scaling Crop Microinsurance: How Will Data & Tech Impact Scale?
This series is produced by Mercy Corps Ventures in partnership with CASE at Duke University. Each article explores the challenges the microinsurance industry faces to scale to support more than 600 million smallholder farmers around the world in becoming financially included and more climate resilient.
Find out more about the project here, and stay tuned for the full series of six articles here.
Our Key Takeaway
Technological advances in the microinsurance industry are necessary to unlock impact at scale — but they’re not enough. On the positive side, data and technology can address key operational challenges such as sign-up and claims processing friction, customer reach, and product design and customization. But the barrier of premium cost for most smallholder farmers (based on increasing risk) is unlikely to be offset by these advances.
Basis risk. Robust indices. Customer insights. Claims processing. Trust. Timely farming advice. Social impact. There are many areas in the creation and provision of crop microinsurance that can be improved with the availability of more precise and timely data and improved technology. While many of our interviewees spoke to the need for better data and innovative technologies to enable the efficient scale of crop microinsurance, we wanted to further explore the problems they are trying to solve along with the implications and continued challenges as data availability and technology evolve.
What are the problems that improved data and technology could help resolve?
Data and technology have the potential to improve both microinsurance product design and distribution, which could address some of the key challenges to impact at scale: meeting customer needs, gaining customer trust, reaching customers, and managing costs (as discussed further in “The Basics: What is Microinsurance?”). On the product design side, stakeholders are eager for technologies and methodologies to:
- Decrease basis risk (i.e., ensure assessments more accurately reflect the experience of the individual farmer).
- Decrease costs of data collection.
“Data is the first line of the budget for the projects we work on.”
— Vance Abissa, CEO of West African microinsurance provider Inclusive Guarantee*
- Understand impact in order to increase and manage it over time.
On the distribution side, our interviewees expressed the need for better data and technology to:
- Reduce enrollment friction.
- Reduce claims processing and payout time.
- Better target customers based on needs.
- Improve wrap-around support and technical assistance to farmers in order to mitigate risks.
How could three key technology advances help enable impact at scale?
Our interviewees spoke about three technological advances that could better enable scale for crop microinsurance: sensing technology and drones, mobile phone penetration, and blockchain. While each provides exciting opportunities to scale more efficiently and effectively, none of these technologies are a panacea.
1. Sensing technology and drones.
While these technologies are already being used in some areas to collect more precise farm-level data, most of our interviewees agree that they are not yet living up to their promise. They expressed hope that as the technologies are refined, the data they provide could help with the creation of more custom insurance products, more accurate assessments for the individual farmer, and more opportunities to monitor and actively support farmers throughout the season. Mark Kahn, Co-Founder and Managing Partner of India-based agritech venture capital firm, Omnivore, spoke to the potential for such data to greatly enable product innovation: “There’s a whole world of more precise granular data that could potentially change the way underwriting occurs. Right now you’ve got spaced out weather stations and satellite data that isn’t very robust. I think in a few years you’ll have on-farm IOT [Internet of Things, such as sensing technology] that can actually throw off hyper-precise parametric data for that farm itself; you’ll also have next gen satellites, such as hyper spectral satellites, that are able to more accurately assess yield than what we have now, which is kind of multi-spectral meets crop cutting. So when those technology leaps occur, it will suddenly get easier to bring some product innovation in, which will be a big help.”
Even with this potential, stakeholders must pay attention to the cost to access these technologies, as well as their fit with the “reality of smallholder farming,” as Pula’s Rose Goslinga described it. Goslinga explained, “We think that there’s definitely a potential for remote sensing data to guide where people go, to determine what kind of samples we have to take. But in the end, we see the future as a hybrid, in that in some locations we will be able to get decent estimates [with remote sensing technology], particularly if yields are good. But the reality of smallholder farming is also very much a reality of intercropping — and intercropping makes yield estimates with satellite data very difficult. Smallholder farming is small pieces of land, and one of the most difficult things in calculating yield estimations through remote sensing is the size of the land. If the yield boundaries are not easily demarcated, the quality of the remote sensing data is not what we would like it to be. It can be fantastic if you have row crops in Iowa, but we’re not talking row crops in Iowa.” Regardless of whether the future is hybrid, or if these technologies play a leading role, their evolution is a trend that should be carefully watched.
2. Mobile phone penetration.
Interviewees see significant opportunities to reach more farmers with insurance offerings as mobile phones are used by higher percentages of the rural population, and as smart phones and mobile money accounts see greater penetration. This penetration could lead to more frictionless sign-up, management and processing of claims, and to the increased use of GPS tracking for both assessments and timely farmer support. However, most interviewees warned that, at least for the foreseeable future, the availability of digital platforms and interfaces in a farmer’s hand will not significantly reduce the need for human touch. For insurtech enterprise OKO, although its model is to leverage mobile to sell crop insurance directly to farmers, Co-Founder and CEO Simon Schwall acknowledges they are “a long way from people being able to self-subscribe.” Schwall sees promise in using mobile apps (such as WhatsApp) for more interactive communication with farmers, while still employing agents in the field and at call centers to build trust and explain insurance to its target customers. One Acre Fund’s Colin Christensen expressed a similar caveat, based on the organization’s 15 years of experience: “In general, with technology, we have not been able to take the field officer out of the equation because farmers are skeptical about engaging with new technology — or lack the digital literacy to master it — outside of a few exceptions.”
While our interviewees acknowledge that many donors and investors see mobile phone penetration and digitization as a game-changer, they also emphasize that the underlying process must be proven robust first. As Acceso COO Rob Johnson noted, “technology is the simple part,” but it’s the operations, process, and rollout that require significant attention and investment.
3. Blockchain technology.
While the many applications of blockchain are still being discovered, some stakeholders have begun applying the technology to create “smart contracts” for the insurance industry (i.e., contracts that are automatically executed when certain conditions are met). Crop insurtech ACRE Africa is currently piloting the use of smart contracts with farmers in Kenya to determine if the technology can help decrease operational costs, streamline and simplify payouts, and automate claim inquiries. It is ACRE Africa’s hope that these changes will create a better customer experience and increase farmers’ trust in the insurance product. Other iterations of smart contracts for insurance also endeavor to create more customized products to better meet client needs (as an example, see Etherisc’s work on blockchain-powered hurricane insurance).
Yet, even if smart contracts live up to their promise, there are still significant challenges around uptake. As mentioned alongside the other advances, the need for human touch in distributing insurance remains, which will continue to drive operational costs even as smart contracts lower some administrative ones.
So much data! Will it be shared? If so, how?
As more data is collected and used by public and private players, how can it be shared to spur market development and impact at scale? Our interviewees shared some reflections and experiences on the potential they see for their impact and bottom line.
Proprietary data and market development: how are stakeholders balancing them?
Annette Detken, InsuResilience Solutions Fund’s Head of Management, shared her thoughts on how data sharing will impact the development of the crop insurance market: “Well, market development depends very much on how proprietary information is handled, who owns the [customer] platforms that are being developed, and who does and does not have access to them. There can be some tension that has to be balanced, of course, because if you want to develop a market, and you have somebody who has proprietary information — especially on these platforms — they can create a monopoly if the access is very limited. But you also need to scale up in order to really reach the point where this is really profitable — so that is somewhat contradicting an open market development. So we are also wondering, to which extent should such a platform be a public good? And who should finance it?”
- The challenge and opportunity in sharing data between strategic partners. Data is needed to build good products, but, as many interviewees shared, it can be a challenge to collect existing data from other stakeholders. Insurtech OKO has worked to actively share its data with partners who can use it to benefit smallholder farmers, such as sharing crop data with a private company working to more accurately map smallholder fields, and sharing repayment rate data with microfinance institutions in order to negotiate better rates for OKO clients. However, OKO has also run into challenges convincing other private enterprises to share their strategic data — likely a result of the time involved for enterprises to make datasets shareable (e.g., time anonymizing, and cleaning, etc) and concerns about propriety. Omnivore’s Kahn validated those challenges and pointed out the role that government could play to shift ownership: “I don’t think agritech [agriculture technology] companies are going to voluntarily share data. I think it’s going to be [shared] when we have public protocols that give farmers ownership of their data that they can share voluntarily, possibly participating in the monetization thereof. I think most of these [agritech] companies see data as the new oil, and the thing they’ve gone out of their way to collect — and trying to get them to share the data is a challenge.”
- The opportunity to monetize data. In addition to using the data it collects to inform its own products and strategies, insurtech Pula has begun to explore earning revenue by selling season’s progress and harvest outcome data to its various segments of customers. If successful, other insurtech companies are likely to explore supplementing their revenue with this same approach, which could make companies more profitable but also further increase the cost of data.
- Data-sharing collaboratives among key stakeholders. The Insurance Development Forum (IDF), a public-private partnership bringing together insurance industry and international development players, was created to “optimise and extend the use of insurance and its related risk management capabilities.” One way in which IDF is achieving this goal is by bringing together private and public sector actors to share data from pilot projects to encourage the creation of more robust and scalable inclusive insurance offerings.
The (time-consuming) reality of leveraging open source data.
If and when more data is available to create and distribute better insurance products, stakeholders will need to pay attention to the time and costs to use it. Insurtech WRMS provides a good example of this burden as it uses many sets of data to understand and model the risk for its products. Examples of the datasets it collects and integrates into its models include yield data from government experiments in different states (collected from government sources and insurance players); open source data from the government including soil, irrigation, and groundwater maps (which WRMS then needs to task a team to digitize); its own remote sensing outputs and weather station data; and collecting samples in villages to correlate the government data. How could these processes be more automated?
Wait. A big caveat on the limits of data & technology advancements.
One Acre Fund’s Global Policy Director, Colin Christensen, expressed a concern (also shared by several other interviewees) about the limitations of these product and distribution innovations in the face of financial constraints. “The big limiting factor for [crop] insurance markets [in developing countries] is financial. You can tinker around the edges and give technical support and actuarial support and incorporate things like remote sensing — but you’re not fundamentally going to change anything unless you push more money into the system, and dramatically increase payout amounts via products that are affordable for farmers.”
Ted Pantone, Co-Founder and CEO of health insurtech Turaco, echoed this sentiment: “I don’t know crop intimately, but the only way to do it efficiently is with good technology. The actuarial science gets much more interesting, the more tech that you overlay. But I don’t think it changes the fundamentals: the economics just don’t make sense for the smallholder farmer to pay to cover that risk.”
So how do we address the issue of who pays and for what? We researched this too, and you can find out in the next article, “Scaling Crop Microinsurance: How Does “Who Pays” Impact Scale?”
This article was written by Kimberly Langsam, CASE at Duke, and Jane Choi, CASE consultant, and released in November 2021.
*Unless otherwise noted, all quotations in the articles are from interviews conducted by Kimberly Langsam and Jane Choi between May and August 2021.
