My job has the great advantage of bringing me in touch with some outstanding and truly brilliant people. As an example, last year I spoke at the PrecisionAg Vision Conference in Phoenix (and I’ll be speaking there again this fall on October 12), and ran into a number of companies that were applying data science and big data principals to the world of agriculture. Having grown up in Charles City, Iowa, I have a special place in my heart for farming and the importance of agriculture to the health and financial success of our country.
One company that I ran into – Aglytix – really impressed me with their application of data science to some of the fundamental decisions that farmers need to make in order to optimize yield while minimizing costs. Aglytix’s approach to first identify the farmer’s most important decisions and then apply data science to optimize those decisions plays right to the heart of the approach that we teach our customers (see Figure 1).
And while I am totally fascinated by Aglytix’s analytic solutions, here is what I was even more impressed with:
- Aglytix is a 15-person company. This is NOT a Fortune 500 company with an over-flowing abundance of business analysts, data engineers and data scientists. 15 people. 15 people with deep subject matter expertise in the area of farming (several came from farming communities like me) who have learned data science from leading data science hubs like North Dakota, Minnesota State, Mankato…not exactly Stanford, Cal or MIT.
- Located in Mankato, Minnesota. That’s 100 miles away from Minneapolis, and 120 miles away from the Big Data Mecca of Charles City, Iowa.
It is exhilarating to see that big data is alive and well in the Heartland of America. In fact, the “democratization of analytics” driven by economics of big data is enabling small communities and small businesses to thrive and compete against much larger and better-financed organizations.
I had the opportunity to spend some time with Jerry Johnson, the founder and CEO of Aglytix. Here is what I learned from our conversation.
Lesson #1: Focus On The Customers’ Most Important Decisions
“Organizations don’t fail at big data because of a lack of opportunities; they fail because they have too many.”
That’s one of my most important messages, and Aglytix has taken it to heart. Aglytix has focused on one use case at a time to build out its suite of lean farming analytics. For example: quantifying the impact of crop stress. Crop stress decreases yield by inhibiting crop growth and decreasing the maximum potential crop yield. Crop stress is caused by factors such as diseases, insects, drought, and salinity and excesses of trace elements in the soil.
Aglytix provides crop stress analytics to help farmers understand the financial impact of crop stress on the farmer’s yield and finances (see Figure 2).
The analytics in Figure 2 show the impact of crop stress over time, both from the perspective of the field (the bottom progression of crop stress on the fields) as well as a quantitative perspective from the chart in the upper right of Figure 1. The types of actions the analytics can recommend to reduce crop stress include the optimal application of water, fertilizers, herbicides and pesticides (both too much and too little can bad for the crops).
Lesson #2: Lead with Recommendations
“Don’t just throw more charts and data at me. Tell me what I need to do!”
This was one of my early lessons from my Yahoo Advertiser Analytics experience. Farmers, like Yahoo Media planners and campaign managers, don’t want “informative” charts and dashboards; they wanted “actionable” insights and recommendations (prescriptive analytics) regarding what they need to do to improve yield and control costs.
Another Aglytix use cases focuses on optimizing weed control. There are numerous variables that a farmer needs to consider in order to optimize weed control from a yield versus cost perspective. And as is typically in any real-world situation, those variables are in constant flux. So instead of trying to turn the farmer into a data scientist by giving them charts and reports, instead Aglytix helps the farmer become a more effective farmer by providing recommendations as to what actions to take (see Figure 3).
The prescriptive analytics in Figure 3 provides the farmer or agronomist with the analysis of what fields to spray, with what herbicides, at what rates and with what pressure in order to optimize weed control. With prescriptive analytics, we don’t need to force the users to guess what to do; we can just tell them what to do with recommendations (think Netflix recommending movies to watch or Amazon recommending products to buy).
Of course the key to successful prescriptive analytics is to measure the effectiveness of the recommendations. In this case, we’d want to know the results of the farmer’s weed spraying campaign so that we can measure how effective it was, and use the results to fine-tune our analytic models.
Lesson #3: Create a Use Case Roadmap
“The best way to guarantee that you don’t solve any problems, is to try to solve them all”
Because of the bounty of opportunities against which to apply data and analytics in order to optimize key operational processes, organizations need to embrace the power of prioritization and focus (see the blog “Big Data Success: Prioritize Important Over Urgent”). The results of prioritizing and focusing will yield a use case roadmap that addresses both cost savings and revenue (yield) improvement, but on a use case-by-use case basis (see Figure 4).
Figure 4 provides a use case roadmap for how farmers can leverage big data to improve yields (increase yields by 40%) and reduces costs (decrease costs by 20%), but does it one a use case-by-use case basis. The big bang effect – spending 5+ years and $15M+ before realizing any financial benefits – have gone the way of ERP implementations.
Instead, organizations have an opportunity to build out their big data architectures, technologies, data science and data lake one ROI-positive use case at a time.
Lesson #4: Make it about the money!
“I don’t care about the 3 V’s of Big Data; show me the 4 M’s of Big Data: Make me more money!”
In the end, big data is about the 4 M’s of Big Data: “Make Me More Money” (see Figure 5).
For example, crop establishment requires farmers to make critical planting decisions including how to plant (e.g., pre-germination, seeding depth, broadcasting) given different soil situations (e.g., wet, dry, soil acidity, soil composition), seed rates, replanting options and pest control.
Figure 6 shows the analytics that not only provides recommendations across the different crop stand variables and decisions, but even shows an estimate as to the financial impact of those decisions.
Now if that’s not a “Make Me More Money” analysis, then I don’t know what is!
Big Data is starting to show up everywhere, and is no longer just the dominion of large organizations. This is the democratization of big data, where organizations of any size can leverage data and analytics to power their business models and win against larger competitors. The big data revolution will truly be in full swing when organizations of all sizes embrace analytics as a business discipline (think “Big Data MBA”), and not just something that IT does for them.
Also, I will be speaking at the PrecisionAg Conference October 12 in Phoenix. I expect to again be meeting some truly brilliant people!