It is worth saying the awkward thing first, because it will save you a month of demos: most growers who go looking for accounting software do not actually need accounting software. Their accountant is perfectly happy. The ledger is fine.
What is not fine is the gap between the system that runs the operation and the system that holds the books — a gap currently bridged by a person, a spreadsheet, and the last week of every month.
That is a different problem with a different answer, so this guide is about the gap rather than about ledgers.
Why growing breaks the standard assumptions
General business accounting assumes you buy or make a thing, it has a cost, you sell it for more, and the difference is margin. Live goods disagree in four specific ways, and every one of them shows up as a reconciliation problem later.
Cost accrues over time rather than at a moment. A tree in its third year has absorbed three years of space, water, labour and overhead. There is no purchase invoice that tells you what it cost, and a costing model that cannot express duration will always be guessing.
Shrink is normal, not exceptional. A percentage of the crop does not make it, and that loss is a real cost carried by the plants that did. Systems built for boxes treat shrink as an incident to be investigated. In growing it is a line in the plan.
The same crop is several products. A #3 and a #15 of one species have different prices, different customers and different footprints. Grade and size are not variants of one SKU in any way that matters commercially — they price separately, they sell separately, and they need to cost separately.
And on retail programs, what you invoice is not what you shipped. Under pay by scan you bill for what sold at the store, which means the number your invoice runs from lives in someone else’s system until it arrives in yours.
Where the re-typing actually happens
When a grower says the accounting is painful, it is almost never the accounting. It is one of four handoffs, and it is worth identifying which one is yours before you look at any software.
- Invoicing — orders exist in the operational system, invoices get created again in the accounting system.
- Credits — a customer disputes or returns, someone raises a credit memo in one system and remembers to reflect it in the other.
- Payments — remittance arrives, and matching it back to what was actually invoiced is manual.
- Costing — margin per product or per customer cannot be seen anywhere without exporting both sides and joining them in Excel.
Most operations have two of these badly and two tolerably. Knowing which two is the single most useful thing you can bring to a vendor conversation, because it turns a feature demo into a conversation about your actual month end.
Replace the ledger, or reach it?
This is the fork, and it is usually decided by whoever speaks first rather than on merit.
Replacing your accounting system means retraining the person who does the books, migrating history, and re-establishing every relationship with your accountant, your bank and your auditor. It is a large project justified only when the ledger itself is genuinely inadequate — which, for most growers, it is not.
Reaching it means the operational system generates the commercial records where the work happens, and pushes them across. The books stay where they are and stop being typed twice. For most operations this is the smaller project and the better answer, and it is what GrowerLive is built to do: pricing and costing live alongside the orders they belong to, farm invoicing runs from orders and from scans, credit memos are tracked against the orders they relate to, and invoices, credit memos and payments received sync across to QuickBooks.
The integration was deliberately built generically rather than around one vendor, so connecting a different accounting system is a configuration exercise rather than a rewrite. If you run something else, that is a conversation rather than a dead end.
The two details that decide whether finance will use it
Both of these sound like footnotes in a demo and turn out to be the whole thing in practice.
The first is approval. Finance teams are, correctly, unwilling to let another system post directly into the ledger unsupervised. An integration that requires a person to approve a batch before anything posts is the difference between one they will switch on and one they will quietly leave off. Ask whether approval is optional per transaction type, because starting with invoices and adding the rest later is how this actually gets adopted.
The second is dates. Your operational calendar and your accounting periods are not the same calendar, and a system that assumes they are will put transactions in the wrong month every time the two disagree. Date mapping is not a nice-to-have; it is the thing that stops month end becoming an argument.
One more worth asking about: what happens when a transaction fails. The right answer is that it retries and then stops, loudly. A sync that keeps retrying indefinitely against a ledger is worse than one that fails.
What to ask a vendor
- Which accounting systems do you sync with today — not on a roadmap, today?
- Which record types sync: invoices, credit memos, payments received? Can I choose?
- Can a person approve a batch before anything posts to my ledger?
- How do you handle the difference between my operational dates and my accounting periods?
- When a transaction fails to post, what happens, and who finds out?
- Can I see margin by product and by customer without exporting to Excel?
- If I run pay by scan, does invoicing run from scans or from shipments?
- What does costing look like for a crop that has been in the ground for three years?
- If I would rather keep part of this manual, can I export it?
That last question is a good character test. A vendor who is comfortable with you exporting to Excel is a vendor who is not relying on lock-in, and there is always some part of the process a grower will want to keep in their own hands.
The honest summary
If your books are wrong, fix your books. If your books are fine but assembling them takes a week of retyping and a spreadsheet nobody else understands, you do not have an accounting problem — you have an integration problem wearing an accounting costume.
Diagnose which of the four handoffs is actually hurting, then ask vendors about that one specifically. It is a much shorter path than evaluating accounting packages you were never going to buy.