Wheat From The Chaff: Distinguishing Valid Farm Data Services

16:57PM Aug 29, 2019
Drone technology on farm

"I'd like to remind everyone that data are intangible and digital, not like tangible goods we're used to, e.g. grain, livestock, equipment, and farmland.," says Terry Griffin.( MGN )

This guest commentary is provided by Terry Griffin, Associate Professor and Cropping Systems Economist, Department of Agricultural Economics at Kansas State University 
[email protected] 
@SpacePlowboy 

While presenting at meetings, I’m often asked my thoughts about specific farm data service providers or more general inquiries into which companies are reputable. Data service providers are the companies that gather farm data to build platforms and usually deliver community analyses. I usually think to myself how I’d like to publicly share my insights, experiences, and observations with the audience but when asked my opinion I skirt the questions with some sort of vague response with the added offering of advice to sit tight, relax, and watch before making any decisions. I wanted to take a moment to explain my reasoning for not directly answering those questions plus offer some suggestions to assist growers in making the best farm management decision for their operation. 

Opinions of data service providers are often made about their current business models and suspected exit strategies; insights often obtained during networking events. Focusing on facts rather than opinions requires more testing that may be infeasible. Unlike products that come in a jug or a bag that universities have routinely tested for many decades in small-plot controlled experiments, testing farm data services has not been as straightforward. Even “consumer reports” style methods haven’t been reliable for assessing the merit of providers. 

Researchers attempting to evaluate data service providers by conducting deliberate controlled experiments likely fail due to the lack of true controls. Since data service offerings typically revolve around community analysis, i.e. data from many other growers, and each provider likely has a different number of growers, i.e. sample sizes, but also potentially a different set of growers in respective systems, it stands to reason that any comparative analytics, e.g. benchmarking, is expected to return a different set of results even for the most reputable providers; therefore testing farm data service providers by merely submitting data from a given test farm would not be sufficient to adequately assess providers in the same way agronomists test a new herbicide.

Given attempts to evaluate farm data services are likely to be inconclusive at best, no incentives exist for providers to maintain quality assurance procedures or even to truthfully report basic information. In practice, a reputable provider and imitator service may appear very similar and could be indistinguishable. Imitator services could easily report how an actual farm performs against a set of made-up data, e.g via simulation by using random number generators; a likely scenario while the data service provider attempts to expand their database platform by attracting additional growers. 

Incentives for farm data service providers are to acquire as many growers and observations as soon as possible before competitors beat them to the finish line. Providers are racing in a winner-takes-all competition. Farmers, on the other hand, are not in a hurry; and this is one of the few advantages they have. Although many farmers feel pressure to immediately join farm data service providers, growers are reminded to leisurely sit back and watch the race play out so that informed decisions can be made regarding which provider benefits your needs rather than trying to ambiguously choose one out of possibly dozens. Joining a farm data service provider is imminent in the long run; however, growers’ decisions should be based on whether the benefits of joining a specific service outweigh perceived costs of doing so. 

As a final word, I'd like to remind everyone that data are intangible and digital, not like tangible goods we're used to, e.g. grain, livestock, equipment, and farmland. A third-party evaluation of farm data service providers will not likely occur, at least in the same way that university researchers test products that come in a jug. Although I would love to publish a rigorous comparison of farm data service providers, a research-based report of that kind is not forthcoming any time soon. To navigate digital service opportunities, growers must treat data for the intangible good it is rather than as the tangible goods that are familiar. 


Dr. Terry Griffin will present three breakout sessions at the 2019 AgTech Expo, Dec. 16 and 17 in Indianapolis: 

  • Dispelling Common Myths of Precision Ag (repeated)

  • Blockchain: The Big Picture and the Nitty Gritty

Register for the 2019 AgTech Expo here