This phrase is usually attributed to W. Edwards Deming, one of the fathers of modern quality management. Whether he actually said it or not, it summarises perfectly what is missing in most berry farms I visit. Not ideas, not products, not willingness to try new things. Data.

Almost every berry company I visit has trial ideas. A new variety, a different way of training the plants, a new pot or substrate, a product applied on the leaf or through the drip. And many of them actually run the trials. But very few collect the results, at least in a way you can actually use.

Testing is how we evolve. It is the closest thing we have to science at farm level. A consultant can bring ideas from other regions, but there are always points that need confirmation, because what works in one part of the world may not work in another. Different climate, different water, different labour. That is why, in many of my visits, I propose trials. Confirmation before scaling.

Here is the frustrating part. Most of the time we “only” need one number at the end: kilos harvested in the trial vs the control. That’s it. And this is exactly where everything falls apart. There is no control. Or we placed three different trials in the same block. Or the trial rows were harvested together with the rest of the field, so the data is simply gone. Months of work, zero conclusions.

Then we have the two classic producer profiles. The first absorbs every novelty on the market, new varieties, new products, new biostimulants, without testing any of them. The second is more isolated and keeps working the way they always have. Both would benefit from exactly the same thing, simple trials with real data. With them, we could prove whether changing variety makes sense, or whether that new foliar product pays for itself.

Without this proof in the field, we get both extremes. Producers applying everything that exists on the market and is “in theory” good, ending up with production costs far too high. And producers frozen in time, still saying that controlling pests in raspberry with beneficial organisms doesn’t work, because they never tested it. Or tested it once, with the wrong protocol, and closed the subject forever.

A useful trial doesn’t need a research station. It needs three things. A control, treated exactly like the rest of the farm. One variable, not five. And separate harvest data, weighed and recorded per treatment, from the first picking to the last. If you cannot harvest it separately, don’t start the trial. You will spend the money anyway and learn nothing.

Trials without data are just expensive opinions. And in berries, with the margins we have today, opinions are a luxury.