I worry that we have created a big problem around farm data

In the AgTech world, we are constantly talking about farm data. We talk about how valuable it is, we talk about what a challenge there is in accessing it. We talk about how worried farmers are about sharing farm data and the different kinds of harm that can result from doing so with bad actors.

All of this talk may have made the problem worse.

Are we creating too much worry?

Could it be the case that we have made farmers more concerned about sharing data than they need to be?

Could it be that we have set expectations that farm data is valuable and that farmers should be paid by other companies who want to access and use their data?

In a world where the factory has no roof is there any point in making farm data such an issue?

Does it matter what we choose to share?

We are quickly approaching a time when vast amounts of data about what happens on-farm will be available to any organization that wants it.

There was a time when the gap between remotely sourced information and directly captured farm-level data was vast. With digitally native agriculture, that gap has narrowed dramatically and will narrow further still.

How much of a gap between the two justifies such deeply held concern over who can and should have access to directly sourced farm-level data.

Are we over-indexing on harm?

I am not suggesting that no harm can come from farm-level data being in the wrong hands or used in ways that are against a farmer’s interest.

But great benefits can also flow from farm-level data being in the right hands. One reason for talking about farm-level data is that there are many benefits that can flow when detailed information is available. Current intense interest in carbon and other ecosystem-service markets are good examples of opportunities that are greatly accelerated by sharing farm-level data.

Participation in these opportunities is also significantly impacted by the unclear (and therefore adoption inhibiting) status of access to farm-level data.

Does it matter how the farm-level data is sourced?

One way to think about it is to analyze the gap in both potential benefits and harms between directly-source and remotely-source farm-level data.

Are there really such big differences based on how the data is sourced?. If a farmer is concerned about the potential negative impacts of sharing more data, do those concerns change much when the data is sourced remotely? There would seem to be no practical way to prevent remotely-sourced data from being acquired and used for any purpose. So is it sensible to worry so much about directly-source data?

If harm can come from access to farm-level data aren’t those harms already live issues?

How accurate is accurate enough?

How much more precise and accurate would remotely sourced data ( and the statistical models the data fuel) need to be before there would be little measurable difference in benefit or harm?

If that gap is already small, then should we worry so much about farm-level data. Is the fear associated with farm data acting as an anchor to adoption for the wrong reason?

Are the great benefits that can be received from real farm-level data being forgone in a futile attempt to avoid the harms?

Choosing to benefit instead of worrying about harm

The more we analyze the possible harms, the better informed we are to have an open and pragmatic conversation about balancing those theoretical harms with more tangible benefits.

We know that it is a natural human psychological response (negativity bias) to over-emphasize negative impacts and underappreciate potential benefits. Farmers are experts in risk management, but has the narrative around farm-level data caused there to be an over-emphasis on the risks associated with sharing data?

The only way to know is to exhaustively catalog the specific kinds of harm that are perceived and analyze their true likelihood and impact. If we can be specific and quantitative about these harms, and also be specific about the degrees to which they relate to remotely- versus directly-sourced data, we can perhaps progress a conversation that for now seems badly stalled.

That constant narrative of fear also slows the adoption of many promising innovations and risks being more harmful than beneficial.