On Farm Experimentation
Evaluating products and practices on the farm has taken many different names, just to name a few:
- On Farm Research
- Field Trials
- Test Plots
Regardless of what we call it, I wonder how much of this potentially meaningful work gets started with the best of intentions but never really gets off the ground or makes a difference to any vested party. Experience tells me the failure rate is high! We can do better by remembering that agribusiness companies and producers have a mutual goal – Learning. Learning how to price a product, whether it works, how to use it, how to increase yields, how to enhance profitability from lower use rates, etc. the list goes on.
We have more tools now than we have ever had before for recording, analyzing, and understanding information generated on the farm, so why is there still such a struggle? Someone has to make all of it happen, put the pieces together, and extract something of value. Software and “artificial intelligence” can’t do this. Remember the scientific method we learned as kids? Relying too much on software or big data platforms leans heavy in the direction of data and results and minimizes the importance of procedures and conclusions. I’m not saying part of the process is more important than the others, but I am arguing that we need to keep the correct perspective.
Procedure – How do we set the experiment up in a way so that shared goals are achieved? How do we get results but let the farmer “farm” and not make experimentation a big pain? How do we reduce error and bias? What parties are responsible for which pieces of information and when?
Conclusions – Company A or Farmer B wants someone (a live human being) to explain within the context what has been learned. A piece of software cannot do this nor will it ever be able to. What did we find out, what is the confidence level, and to what extent can we apply that knowledge to future crops, other geographies, other technologies?