A harvest spreadsheet layout that mostly works

If you're going to run harvest on a spreadsheet, set it up properly before the first trailer moves: one row per load, a fixed set of columns, a second sheet holding field areas, and a pivot table doing the totals. Built that way, a spreadsheet will get you through a season. Built on the fly in the second week of wheat, it won't.

Here's a layout that has survived contact with real harvests, the habits that keep it alive, and a straight answer on where it stops coping.

The load sheet: one row per trailer

Every load gets a row. These columns, in this order, and nothing clever:

Column	What goes in it
Date	The day the load was weighed
Time	Roughly is fine — it settles arguments later
Trailer	Which trailer or lorry, because the tare depends on it
Gross (kg)	The loaded weight
Tare (kg)	The empty weight, taken this week, not in May
Net (t)	A formula: gross minus tare, divided by 1,000
Crop	From a pick list
Variety	From a pick list — it matters when one shed is milling spec
Field	From a pick list, spelled exactly one way
Shed	The destination store, from a pick list
Moisture %	If sampled; blank is better than guessed
Initials	Whoever recorded it, so there's someone to ask

That's the same information a complete weighbridge ticket carries, which is no accident. The sheet is the ticket book, typed.

Net weight is always a formula, never a typed number. The moment someone keys a net weight in directly, you've lost the ability to check it against the gross and tare — and sooner or later you'll want to.

The second sheet: fields and areas

One row per field: field name (matching the pick list exactly), crop, variety, and cropped area in hectares. Check those areas against your RPA maps before harvest, not after — a field entered at 12.4 ha that actually has 10.9 ha of crop in it will flatter its yield by roughly 14% and quietly mislead next year's seed and fertiliser sums.

This sheet is what turns tonnes into yields. Without it you have totals. With it you have tonnes per hectare per field, which is the number actually worth having in November.

The pivot table that earns its keep

Build three pivot tables off the load sheet and refresh them whenever the file is opened:

• Rows: field. Values: sum of net tonnes. Divide by the area sheet and you have field yields.
• Rows: shed. Values: sum of net tonnes. That's shed stock — as of the last entry, at least.
• Rows: crop, then variety. Values: sum of net tonnes. Season totals, split the way the merchant thinks about them.

If pivot tables are unfamiliar, ten minutes with the spreadsheet's help pages covers it. It's the one feature that stops the file being a long list nobody reads.

The disciplines that keep it alive

A layout is the easy half. The habits are what decide whether the file is a record or an archaeology project:

1. Enter loads the same day. Not "when I get a minute". The numbers worth checking each evening only exist if the day's loads are in before you look.
2. Pick lists for crop, field and shed. Use data validation so those columns can only contain values from a list you set up in July. Free typing is where "Long Acre", "long acre" and "L Acre" come from, and each one gets its own line in the pivot.
3. Protect the formula columns. Lock the net weight column and the yield sums so a tired thumb on a Sunday night can't overwrite them with a number.
4. Refresh tares weekly. Mud, rain and a toolbox left in the trailer all move the empty weight. A stale tare puts the same error into every load recorded against it.
5. One file, in one shared place. A copy on the office PC and another on a laptop will disagree by August. Put it somewhere everyone opens the same file.

None of that is difficult. All of it gets skipped in a catchy week, which is rather the point of the next section.

Where it honestly stops coping

Three limits are built in, and no amount of formatting fixes them.

One person owns it. The combine operator can't see it, the cart drivers can't add to it, and if the owner spends Saturday on the drier, forty loads sit in the ticket book unentered. Come Monday, the milling wheat question gets answered from memory.

Entry happens in the evening. The loads arrive every twenty-odd minutes for twelve hours; the typing happens at 9pm. Fields get guessed, tares get invented, and every day of lag makes the backlog worse. With milling premiums commonly £10-£30/t over feed depending on the season, one wrong shed entry that lets 200 tonnes get blended is a four-figure mistake the spreadsheet only reveals in November.

There are no live totals. The pivot is right as of the last entry, which during harvest means it's right as of yesterday, or the day before. Shed-space decisions and lorry bookings get made on squints instead.

If those three bite on your farm — and on most farms over a few hundred tonnes they do — the fix isn't a cleverer spreadsheet, it's recording loads at the moment they exist. Purpose-built tools, Harvestt among them, exist precisely because of those three limits: anyone can record a load from the pad on a phone, and the totals build themselves. But if your harvest is a few fields and one pair of hands, the layout above, entered same-day, is genuinely fine.

Frequently Asked Questions

What columns should a harvest spreadsheet have?

One row per load with: date, time, trailer, gross weight, tare weight, net weight (as a formula), crop, variety, field, destination shed, moisture if sampled, and the recorder's initials. Keep field, crop and shed as pick lists so the names stay consistent, and hold field areas on a separate sheet.

How do you work out field yields from a harvest spreadsheet?

Sum the net tonnes per field with a pivot table, then divide by the cropped area held on a second sheet. The result is only as good as the area figure, so check areas against your RPA maps before harvest rather than trusting an old number.

Why use pick lists instead of typing field names?

Because three spellings of the same field become three separate lines in every total, and untangling them in November is miserable. Data validation takes minutes to set up and removes the problem entirely.

When does a spreadsheet stop being enough?

Roughly when more than one person needs to record or see the numbers during the day, or when evening data entry starts lagging behind the trailers. At that point the file is accurate in November and useless in August, which is backwards — August is when the decisions get made.