Growers characterized some aspects of government regulation to be outof-date with current technological trends and emphasized that government intervention has not played an active role in expediting precision weeding technologies. Though both VCs and startups called out economic development funds and locally-supported R&D tax credits, both stakeholder groups did not consider these government interactions to be as effective as industry-led programs. Another theme that emerged regarding the second objective is that the ideal role of government should be more hands-off because companies should find success in the free market. Some did bring up the benefits of clearly defined funding opportunities to reduce the friction of growers seeking government funding. A study using financial datasets of over 32,000 companies found that subsidies were only effective for short-term innovation while tax credits were favorable on both short and long terms . Subsidies are direct fiscal measures where the government is the project decision maker whereas tax credits are indirect fiscal measures whereby companies can choose their own projects and the direction/purpose of the innovation activity . However, all interviewees agreed that tax credits and subsidies should not be the primary driver or business case for precision weeding adoption. Government support should only fill in initial financing gaps as startups works towards achieving economies of scale.Interviewees from all stakeholder groups brought up the grower preference for equipment ownership, vertical grow racks going against service-based business models. Some precision weeding startups are using or have tried in the past to employ the weeding-as-a-service, recurring revenue model, whereby customers pay the startups for each instance of the service of weeding.
Owning equipment is preferred, especially accounting for the time sensitivity of some agricultural operations. Additionally, government loans and incentives, such as USDA Farm Service Agency Direct Farm Ownership Loans, help growers enter long-term leases or purchase their own equipment. Because of this strong equipment-owning preference, grower skepticism towards a weeding-as-a-service business model can hinder the growth of precision weeding startups. However, literature shows that recurring revenue models, as opposed to the direct sales model, have the advantages of significantly lower capital costs and lower payments compared to traditional loan structures, especially considering that weeding equipment remains idle for the majority of the year . Furthermore, subscription models give growers access to the newest and latest technologies, sharply decreasing the replacement cycle periods . These two revenue models are also not binary: recurring revenue models can derisk the use of new precision weeding innovations, with an initial subscription service proving the value of the technology, before growers transition to purchasing the equipment at the end of the contracting period . Beyond one-time purchasing and recurring revenue models, shared equipment pools managed by grower cooperatives and/or community organizers can derisk growers trying expensive technologies. An expansion of informal peer-to-peer networks, cooperatives can own equipment on behalf of their members and operate at cost. Non-profit or community-based organizations and equipment sharing businesses can also provide such services . Though growers typically prefer equipment ownership, case studies on equipment sharing arrangements, particularly in Europe, have assuaged timing-related fears. For example, a Swedish cooperative used 20 years of data to calculate the economic losses from performing a field operation at a sub-optimal time.
Their results found that the cost savings from shared equipment still outweighed ‘timeliness costs’ . In addition to the non-purchase option that cooperatives provide, contractor services can also be a mode for equipment deployment. The contractor does not necessarily have to be the manufacturer . Instead, equipment dealers or other businesses can train third party technicians and offer contractor services. In fact, Bavarian small scale farmers preferred both contractor services and shared equipment models over equipment ownership and working directly with the manufacturer . Despite such literature proving a preference for cooperative sharing pools, the interviewees of our study did not express similar views.The third objective regarding the user journey of growers adopting precision weeding technologies examined grower and startup concerns and touch points from awareness to continued use and advocacy. Both startups and growers emphasized in-person networks as the most important touch point and platform by which growers learn about new technologies. The influence of peer-to-peer networks brings up the question of perceived social risks: studies have shown that farmers who believe in combating climate change in their operations are concerned about the risk of negative social perceptions and thus reduced access to peer-to-peer networks . In this study, many Northern California farmers implemented climate adaptation farm management practices, such as water-effective irrigation. However, many remained sensitive about the negative political/social framing of climate change, strongly justifying their environmental adaptive actions with co-benefits, such as economic efficiency . Additionally, in a pollination management study in Michigan, researchers uncovered several large advice networks in which growers can reach other growers and their partners in a maximum of two to three hops and that 26 % of all communication mapped was from grower-to grower . In our study, growers cited word of-mouth from their peers as one of the key touch points influencing their journeys through the awareness, consideration, and purchasing phases.
Because many startups lack direct connections with growers, growers may look towards their peers to vet potential startup partners. In addition, aligning with Petersen-Rockney , the competitive landscape of specialty crops lend itself to the high stakes of intra-stake holder dynamics and social perceptions.For the first objective about the compatible motivations between precision weeding stakeholders, we found that the most common motivator was labor concerns, which was cited by 13 out of 17 interviewees. Some stakeholders added additional details about the labor pressures of organic farming in California, the competitive labor market, the hiring budget difficulties because of California’s increasing minimum wage, and the role of precision weeding technologies in increasing the efficiency of labor. Other commonly cited motivations were costs and precision weeding’s potential to ‘transform agriculture.’ Answering the second objective of collaborative models between stakeholders and their limitations, we found that the grower-identified blockers to adoption included competition between growers, oldschool mentalities such as equipment ownership, and a lack of connection between startups and growers. Additionally, startups and VCs also mentioned and expanded upon the growers’ urge to own their own equipment, coming into conflict with the weeding-as-a-service business model that some startups have ventured into. The startups agreed with the limitation that some startups lack connections to growers and suggested a few solutions: hiring employees who have worked in the agricultural industry and have local connections and startups developing strategic alliances with committees of growers. Both startups and VCs discussed their symbiotic relationship: VCs provide startups with portfolio support to accelerate business growth and mature company governance in exchange for equity. Exclusive to the agricultural industry, some VCs spoke to their role in providing startups on hands-on, agriculture-related support such as building a grower advisory board. All three stakeholder groups had varying opinions of the role of government in the precision weeding ecosystem. While all agreed that government-funded university research and the UC Cooperative Extension were helpful, pipp racking growers and startups viewed all tiers of government as being out-of-touch with grower needs and not providing enough R&D support to accelerate the adoption of precision weeding. The growers’ distrust of government is consistent with existing research. Regarding the third objective about the grower user journey, the most common touch point identified in the growers’ awareness phase was in-person networking, consistent with existing research about the influence of peer-to-peer networks in spreading information. Most of the areas of improvement were in the consideration/piloting phase, such as educating growers about misperceptions towards weeding-as-a-service business models, such as lack of equipment availability. As for the study’s limitations, the experimental design purposely was exclusively semi-guided qualitative interviews, as opposed to a mix between interviews and surveys, to examine in-depth perspectives and experiences. A limitation inherent in this experimental design was a lack of breadth, as the stakeholders interviewed in this study did not thoroughly represent the entirety of California’s precision weeding ecosystem, especially considering our focus on only three stakeholder groups. Future research could expand beyond our interview approach to include willingness-to-pay surveys to ascertain grower attitudes towards precision weeding. In addition, future research can broaden our research objective of compatible motivations to develop joint value propositions.
Better aligned value propositions will help startups communicate with growers and may influence startup R&D direction to better suit grower needs. Another limitation we faced with the interviews is that the agtech investing space, particularly for investors who have interests in California agriculture, is quite small. Because ‘all VCs know each other,’ some venture capitalists we interviewed or reached out to interview had concerns about anonymity and/or having diverse-enough viewpoints. Additionally, government outreach proved difficult with many potential interviewees canceling after learning about the informed consent conditions. However, this limitation was not a significant hindrance as the three core stakeholder groups did not include the government. This does raise a potential future research direction that more directly examines agriculture-related government endeavors and public-private collaborations. A limitation in an aspect of our data collection method was concerns regarding interview standardization. Though we followed a guide of pre-prepared questions for each objective, some interviewees cut the interview short while some were very generous with their time. The shortest interview was twenty-three minutes while the longest one was over an hour and a half. Therefore, in our results, we only featured stakeholder group maps, not individual maps. The broader implications of the study are for the governmental role of encouraging new technologies and the usage of cognitive mapping for agtech ecosystems. Precision weeding technologies respond to grower needs with conventional weed management. Resulting from labor pressures, the rising costs of chemical inputs, and market and government pushes towards environmental sustainability, California growers are pursuing alternative techniques such as buying machinery from new precision weeding startups. However, the adoption of precision weeding technologies is at times hindered by negative growers’ perceptions of startups and new technologies. In addition, despite strong public sector involvement in other digital agriculture sectors, this study found relatively low government involvement in precision weeding. Therefore, the policy implications and our recommendations are less on the regulatory side, but more so for governments to have a soft role in shaping standards and establishing legitimacy for new technologies. The cognitive mapping presented in this study can be applied to other emerging agtech ecosystems. Because most venture capital firms that invest in precision weeding also invest in other agtech startups or industry-agnostic hardware/deeptech startups, the interactions between agtech startups and venture capital firms may remain similar. However, future research may find varied interactions between the startups and growers. A future research direction may apply the same stakeholder mapping/qualitative interviewing methodology to other emerging agrifood technologies in regions outside of California. Like precision weeding technologies, crop harvesting robotics for specialty crops have achieved impressive prototypes in academic settings . Because of the similar high hardware costs and relatively low on-field use throughout a season, single-purpose autonomous robots are a suitable candidate for this study’s methodology. A study examining autonomous robots for soft fruit cultivation in the UK discovered that labor shortages are driving farmers to become interested in and optimistic about the potential of autonomous robots in picking, disease treatment, and harvesting . While our results found that growers were cognizant of the limitations and lower technology readiness levels of the startups’ initial prototypes and were sometimes willing to adjust their expectations accordingly, some studies have shown a deeper involvement between these two stakeholder groups at this early demonstration/piloting stage. For example, a project in New Zealand had been researching collaborative design efforts for a robotic apple harvester. Because orchards must be ‘robot-ready,’ such as having denser rows with trees lying flat against trellises, researchers found that some orchardists have been adjusting their landscapes accordingly to accommodate future technology adoption . Beyond leafy greens and fruit orchards, site-specific management is ideal for the heterogeneous nature of vineyards and our study’s methodology can be applied to precision viticulture. As the technology and user readiness of precision agriculture hardware develops, stakeholder mapping canidentify gaps in the existing ecosystem and offer both public and private entities opportunities to develop mechanisms that accelerate technology adoption.