Agriculture is lifting off in the world of start-ups. Google’s Eric Schmidt recently announced an accelerator dedicated to backing startups in the domain. This is an encouraging development that could bring agriculture to the forefront of the digital tech ecosystem, and might even give it a top position in the field of tech innovation.

But despite the opportunity of commercial venture capital directing itself to agriculture, there is a major overlooked problem that tech development in ag faces, and that is the condition of its ecosystem around entrepreneurship and startups.

The agricultural sector suffers from a backlog in terms of informed entrepreneurship skills. Generally speaking, agricultural engineers, who either design solutions for farmers, or are farmers themselves, know very little about entrepreneurship. They build things animals and plants like, but not necessarily their human users. Agriculture knows tons about engineering, but little about the innovation that is required to successfully support new technology adoption in the farmers’ market.

Leadership is the second ecosystem problem in agriculture. In a dynamic landscape, like that of agriculture at the moment, it’s important that there are leaders out there that can define an end to which the game will likely play out. From defining such an end, you can then work backwards, and make the hard decision about what activities and initiatives you should be undertaking now, to move in a trajectory towards that end, even if this goes against the grain of conventional wisdom. (I strongly encourage you to read John Hagel’s recent two posts on the future shape of strategy)

However, many of agriculture’s leaders don’t give much for landscape thinking and futurism. They prefer to work with linear progression, departing from the handful of business models that have ruled agriculture for the past 50 years, for their predictions of the future. This is not such a fruitful perspective for a market environment in which technology tends to blur, rather than affirm classic industry boundaries.

The ecosystem challenges are not only of the broad landscape definition kind. They also lie in the specificities of tech design for farmers. Particularly the basic technology user experience for farmers has not been understood thus far. For instance, a startup called Farmobile is betting on its own hardware module to function as a data bridge between reading out data from tractor and machine sensors, and mass storage on the cloud. Farmobile explicitly states that they don’t work with mobile phones and tablets, because they break, get lost, have batteries that drain too quickly, and are cumbersome to use in pairing for farmers and their farms hands.

Convention would dictate that the mobile platform be used. Such a choice for explicit distantiation as Farmobile has taken from mobile would generally be considered as Silicon sacrilege. But I am convinced Farmobile knows their users better than convention dictates, and is making the right bet on UX. A bold decision which nobody, or no precursor could support them to make as of yet.

The bottom line is that there is no ecosystem yet of founding teams with experience and mentors and investors alike, who understand what lies ahead for agriculture and the practical challenges to overcome. That ecosystem is yet to be built. I predict, nay warn, that if “ecosystem” remains optimistically overlooked in the new investment strategies that are popping up for agriculture, that the lack of entrepreneurship, leadership, and design is going to be one of the big, hard walls that tech development in agriculture will hit.