Ask most farms what a crop yielded this year and you will get a single number — total bushels, total pounds, total bins — for the whole operation. It is the easiest figure to produce, because it usually comes straight off a scale ticket or a settlement sheet that already adds everything together. It is also the number that hides the one thing a yield record is actually supposed to reveal.
A farm-level yield is an average across every block that grew the crop, and an average, by construction, erases the difference between the blocks that made it and the block that dragged it down. If four blocks each performed close to expectation and a fifth performed poorly, the farm-level number can still land somewhere respectable — and the fifth block, the one place a decision could actually change next year’s outcome, never gets flagged.
A single number invites the wrong conversation
When a season is summarized as one figure, the conversation that follows tends to be about the weather, the market, or the year in general — explanations that apply equally to every block and therefore explain nothing about why one block underperformed the rest. A block-level record forces a more useful conversation, because it isolates the variable that actually changed: this field, this input, this variety, against a backdrop where the weather and the market were held constant across the whole farm.
An average is a number with the useful part removed
The whole value of a yield record is supposed to be diagnostic — it should tell you which fields are working and which are not, so the next season’s decisions about seed, inputs, or even whether to keep farming a given block are grounded in evidence rather than impression. A single combined number cannot do any of that. It answers “how did the year go” and says nothing about “where did the year go wrong.”
This is not a subtle statistical point — it is the difference between a record that can guide a decision and one that can only summarize a season after the fact. If the goal is the former, yield has to be attributed to the block it came from, not blended with every other block that happened to grow the same crop.
Attribution requires area, not just quantity
A per-block yield record is only useful if it is normalized by area. A block that produced less in total than another block is not necessarily underperforming — it may simply be smaller. The question that matters is yield per acre, or per hectare, which requires knowing both the harvested quantity and the block’s actual area, not an approximate one.
This is why a stable field boundary held as real data, with a known and current area, matters for yield in particular. A block whose boundary has shifted since it was last measured — split, merged, or simply eyeballed — produces a yield-per-acre figure that is wrong in proportion to how far off the area is, which quietly corrupts every comparison built on it.
A yield record is only comparable across years if it is attributed to the same field identity each time, the same discipline covered in crop rotation records. If a block’s name or boundary shifts between seasons — split, renamed, or merged with a neighbor — a yield trend that looks like a real decline might actually be an artifact of comparing two different pieces of ground under one label. Fix the field identity first; the yield comparison is only as reliable as the thing it is being compared across.
The harvest lot is where block-level yield gets lost
Even farms that plant and manage blocks separately often lose the block distinction at harvest, because multiple blocks get combined into one bin, one truck, or one settlement ticket. Once that happens, the yield figure that comes back from the elevator or the packer is already a blend, and no amount of careful farm-level bookkeeping can un-blend it after the fact.
The fix has to happen at harvest, in the same way lot coding has to happen at harvest: keep each block’s harvest weighed or measured separately, even if it is combined into a shared bin or lot afterward for sale. The sale can be combined. The yield record cannot be, or the block-level question this whole record exists to answer becomes unanswerable the moment the truck leaves the field.
A losing block is information, not an embarrassment
There is a real temptation to let a bad block’s numbers quietly disappear into a farm-level average, because a block-by-block record makes underperformance visible in a way a combined number does not. That instinct runs exactly opposite to the record’s purpose. A block that consistently yields below the rest of the farm is not a fact to obscure — it is the single most actionable line in the whole record, the one that tells you where a change in variety, input, or even continued use of that ground would pay off.
Measure it, keep it, and look at it plainly. A yield record that only shows the farm doing fine on average is not protecting anyone from bad news; it is just delaying the moment the bad news gets noticed, usually until it has repeated for several years.
It is also worth resisting the opposite temptation — writing off a single bad year at a single block as noise before checking. One poor season can be weather. Three in a row, on the same ground, is a pattern, and the only way to tell the difference is to have kept the block-level number long enough to compare it against itself, the same multi-year discipline behind a rotation record.
Attribute it or the measurement isn't worth taking
If yield is only ever going to be looked at as one farm-level number, the effort of tracking it precisely is largely wasted — the number that results answers a question nobody was really asking. The record only earns its keep if it is broken out by the block it came from, normalized by area, and looked at block-against-block, year over year.
Farm40 ties a harvest record back to the planting and the field it came from, so a yield figure can be exported per block rather than only as a farm total. The limit: an export is a view of what you entered, at the level you entered it — if two blocks’ harvests were weighed together and logged as one entry, Farm40 cannot separate them after the fact any more than a scale ticket could. The separation has to happen at the scale, not in the export.