Charlie Andersen (MBA 2014) grew up on a 200-acre organic family farm west of Philadelphia that produced everything from apples and blueberries to sweet corn. Andersen absolutely loved it—for the most part: “What I disliked were all the little tasks we had to get out of the air-conditioned tractor cab to do,” he says.

Hand-worked chores account for a huge portion of labor on farms, says Andersen. Take, for example, table grapes, which are picked exclusively by hand: California farmers gross roughly $2 billion a year; of that, somewhere between 40 and 50 percent is spent on labor. Those factors inspired Andersen to found Augean Robotics (AGR), makers of an autonomous farm robot called Burro, which has the carrying capacity of a sure-footed donkey and the smarts of a deep-learning computer that allow it to visually distinguish between weeds and crops.

For its initial handful of customers, AGR has focused on high-value, high-margin but labor-intensive crops like berries and table grapes, where Burro can use computer vision to navigate between rows, following pickers to shoulder the haul. The Philadelphia-based company closed $1.5 million in seed funding last March and is ready to scale. “We’ve gone from having a couple of robots in the field to showcase their autonomy to having a fleet of them running miles every day,” Andersen says. And it’s just the beginning of Burro’s potential: As AI capabilities expand, Burro could eventually be adapted to tackle more complex tasks like pruning and picking.

For now, AGR is focused on nailing the initial market fit, Andersen says, but at some point he’ll have to face the question of empire building. Other companies have expressed interest in the data Burro is collecting out in the fields, which could be deployed to build a host of complementary AI tools—like a picking arm that could be bolted to Burro. But weighing the potential value of its data against the resources required to build value from it, Andersen questions how or whether to collaborate. Think of it in Microsoft versus Apple terms: Should AGR adopt a Microsoft Windows–style ecosystem, where it focuses on its core platform (Burro) but shares data with partners that could build out additional tools? Or should it take the Apple approach of owning all the hardware and software and not allowing others to build on it? The sole ownership model made a lot of sense for Apple early on, Andersen notes, but the collaborative Windows approach enabled Microsoft to become bigger, quicker. “Within AGR there’s this notion that you don’t want to share a ton of this data, but there’s also a recognition that these problems are hard to solve and take a lot of resources,” he says.

I have a classmate who had an incredible idea and could’ve started a platform to create cash flow. But the bigger he thought, the more monstrous the idea became, and the more money he needed to raise. It has now become too big to fund, with no proof of concept or track record of success—and the competitors he sought to disrupt are already sweeping over him. Hubris gets in the way of hitting singles and doubles; you hit the home runs when you have already swung at more than one pitch.
—John Tilley (OPM 53, 2019)

There is virtually no way to manage the process and the funding required to do it all yourself. Either you’ll end up giving the business away to an investor to help you pay to do everything now, or you’ll likely be so slow to respond that someone else will be able to see the opportunity and displace you quickly. I’ve been in the venture world for 20 years or so on the entrepreneurial side, and I’ve seen the second, third, or even fourth to market wind up in a much better position to take a leadership role, because the market is more proven by then. Don’t let them do that. Partner wherever you can. You can only do so much as a startup.
—Glenn Lovelace (MBA 1987)

At this stage, the goal should be doing the one thing that is distinctive and value-added very well, and scaling it up to where it is hard for others to duplicate. (Please just be ethical and socially responsible along the way.) It is difficult to make a concrete recommendation about where to draw the line between the closed and the open parts of an ecosystem like this one, but fundamentally, it is not an either-or proposition. Management should set smart criteria for distinguishing between partners (defined broadly) and types of partners that bring appropriate contributions to the ecosystem and those that will be tempted to eat the data for lunch for themselves once they gain access.
—Alex Evans (MBA 1993)

At an early stage, AGR should follow the Apple ecosystem and work on both the hardware and software. This will allow you to understand requirements of modification and research to enhance the product. At a more mature stage or later on, to prolong the life cycle of the product, AGR could switch to a Microsoft ecosystem to focus on product only and let others develop additional tools.
—Vineet Mittal (OPM 37, 2008)

Burro is an innovative product that has a potential for large cost savings, yet it is not protected from competitive developments. Hence, reaching critical mass and name recognition should be foremost in its strategy. The Microsoft approach would enable the company to benefit from the “wisdom of the crowds,” thereby leveraging its core into both market expansion and product enhancements.
—Jean-Louis Danis (MBA 1974)

Got a case? To take part in a future “Case Study,” send an outline of your company’s challenge to bulletin@hbs.edu

Post a Comment