How predictive analytics can help take your farm to the big leagues
There’s an old question in the agriculture industry: You have two farms of the same size growing the same crops only a stone’s throw away from one another. Both farmers planted their crops at roughly the same time, use roughly the same amount of water and fertilizer, and experience the same weather throughout the growing season. They finish their harvest within days of one another, yet after all the dust has settled one operation yields an additional $50 per acre. What made one farm so much more productive?
An industry ripe for change
Much has changed since the dawn of the industrial revolution—from the introduction of the internal combustion engine to drought and pest resistant seeds—but the principles of sowing, growing, and harvesting crops has been steadfast for the last 10,000 years. Plant your field at the right time. Make sure it gets the right mix of water, fertilizer and sunlight. Wait until the last possible moment to harvest.
For all the quantum leaps in weather forecasting, high quality fertilizers, and state of the art combines, wisdom, intuition, and a little bit of luck are still the screws that turn an operation. The right decisions at the right time are the difference between a bumper year or going bust. Misjudging the last or first frost by only a handful of days can be costly.
Too many solutions in search of a problem
Countless technologies have come and gone over the years promising to eliminate the guesswork and help farmers make better decisions. Yet, by and large, most have failed to live up to their bold promises.
Onboard computers, drones, and state of the art machinery are all fantastic when they’re functioning as they’re supposed to. But who has time to learn how to operate these new devices, and what happens when they inevitably break down mid-spray in the middle of the field? No farmer worth his or her salt is going to lose an hour—much less a day—talking to tech support trying to fix a computer glitch that otherwise isn’t getting in the way of caring for their crops.
Some innovations have managed to stick around and make life easier. Many are showing real promise for the future and for the broader supply chain. But in an industry where pennies spent often accumulate faster than pennies earned, highly consequential decisions still boil down to the human factor more often than not.
That’s not to say a lot of promising agriculture technologies aren’t exciting or can’t serve a meaningful purpose. They’re not taking root because they’re attempting to solve farmers’ problems by adding to their workload—when what farmers really need is something that can help make use of the work they’re already doing.
Has agriculture come full circle?
Data analytics is perhaps the worst-kept secret in the business world today. People have been shouting from the rooftops that ‘data is the new oil’ for years, while most of the global 500 have been building empires off predictive analytics to put ads and products in front of the most likely buyers. Yet the agriculture industry has failed thus far to fully capitalize on its potential. Given that oil was once the new agriculture and crops were the original commodity that drove empires, perhaps now’s the time to close that loop.
Let’s return to our conundrum of seemingly identical farms seeing remarkably different results. Conventional wisdom says yields are driven by three factors: inputs, soil, and weather. Follow that logic to its natural conclusion and one can only assume these two farms must not be so similar after all. They’re across the township road from one another, so it can’t be the weather. So, it must boil down to subtle differences in either the soil or the inputs, right?
MNP’s Agriculture and Technology Solutions teams are demonstrating how predictive algorithms can help us find that very answer. We’ve isolated more than 150 operations variables which are consistent across every farm and readily accessible to every producer—and narrowed those down to just a handful of critical factors which consistently predict which operations will be the most productive. So far in our 50-farm test group we’ve not only been able to demonstrate a clear return on investment, but a quantifiable difference of 12 bushels per acre between the least and most productive farms.
Do what you’re already doing—only better
Few operations have the time to set up, learn, and manage their own data management system. In the near future, a data science team will be able to advise the farmer on what needs adjusting, by how much, and the anticipated return (in bushels per acre) based on current market price, all from the day-to-day insights farmers are already collecting.
The best part of this system for producers is unlike other technologies which rely on exclusivity and rapidly become worn out or obsolete over time, this process will deliver better results as it matures and the number of operations using it continues to grow. The larger our datasets, the more accurate our predictions will be, and the more precise calculations farmers can make in the field.
Conventional wisdom says the earlier a crop is planted the larger it will grow. But nobody could ever say for sure how much earlier and how much bigger the yield would be. Finally, we will have some deeply rooted insight to support generations of cultivated intuition.
Shea Ferster is a multi-generation farmer and Business Advisor with MNP’s Agriculture practice in Saskatoon. To learn more about MNP’s Agriculture Operations Improvement service, contact Shea at 306.664.8318 or firstname.lastname@example.org