Seed Production and Conditioning Software: Keeping Certification, Germination and Royalty Attached Through Every Blend | Digital Heroes

The lot that quietly lost its identity

A conditioning plant runs a certified soybean lot in November. Screen size splits produce a main fraction and an undersize fraction. The undersize gets held. In January a customer orders more than the main lot covers, so the plant blends carryover from last year, which was the same variety from a different grower field and a different certification class. Someone rebags. The new bags carry a tag with a germination figure from a test done in October.

Every one of those steps is normal. Together they have destroyed three things at once: the certification class of the blended lot, the validity of the germination figure on the tag, and the royalty accrual, because the carryover was already accrued last season and now the units are being sold again under a new lot number. Nothing was done dishonestly. The identity chain simply had no system holding it.

That is the defining problem of a seed business. A bushel of commodity grain is fungible and its identity does not matter. A unit of seed is worth many times more precisely because of the identity attached to it: this variety, this class, this germination, this treatment, this licence. Lose the attachment and you have converted expensive seed into cheap grain, on paper and sometimes legally.

Why standard software cannot hold a seed lot

There is no established product for this, and the reason is that a seed operation is three businesses stacked on one lot.

Grain accounting software treats inventory as fungible bushels in a bin with a position and a price. It has no notion of a lot that cannot be commingled without consequence, a class hierarchy, or a test result with an expiry.

Manufacturing ERP (Enterprise Resource Planning) handles bills of material and lot traceability, but its model is discrete assembly: components go in, a product comes out, and the genealogy is a tree. Seed conditioning is not a tree, it is a graph with splits, recombinations, carryover across years and rebagging that changes the unit of measure. A lot splits into three size fractions, two of which recombine with material from a different parent lot the following season. Standard lot traceability handles the merge poorly and the year boundary worse.

Agricultural retail software is built to sell seed, not produce it. It knows units, varieties and grower accounts, and knows nothing about field inspection, isolation distance, conditioning yield or certification agency reporting.

So the identity chain lives in a plant notebook, a certification file cabinet and one production manager memory. That works until the person changes or a lot is questioned.

The unit of measure makes it worse. Material arrives as bulk pounds off a truck, is conditioned by weight, is sold to a farmer as units of a set kernel count for corn or as bags for soybean, and leaves the yard in seed boxes and totes that get returned and refilled. Every one of those conversions is a place where inventory quietly drifts away from reality, and drift in a business where the material is worth many times its grain equivalent is not a rounding issue, it is missing money.

What a custom seed production build has to include

• A lot object with genealogy that supports splits, merges and carryover across crop years, where each lot carries its class, its parent lots, its production field and grower, and its accumulated test history. This graph is the spine of the whole system and if a developer models it as a simple parent child tree, the project will fail at the first blend.
• Contracted production: acres by grower, by variety, by class, with the contract terms, the expected production and the settlement basis, so that harvest lots attach to a contract rather than appearing from nowhere.
• Field inspection and isolation records from the certifying agency, attached to the production field and therefore to every lot descended from it. When a field fails inspection, the system should immediately show every downstream lot affected.
• Conditioning operations with real yields: cleaning, sizing, gravity, treating and bagging, each consuming and producing lots with a recorded yield loss. Conditioning yield is a genuine cost driver most operations only estimate.
• Test management for germination, purity, and where relevant vigour and pathogen testing, with retest scheduling driven by label age rules. Labelling limits how old a germination test may be, and the exact window varies by state and by interstate shipment, so the system should hold the rule per destination and warn before a lot becomes unlabelable rather than after.
• Certification tag issuance and agency reporting, with class carried correctly through every operation and a hard rule that a blend cannot claim a higher class than its lowest input.
• Treatment records: product, rate, applicator, date, and the fact that treated seed carries handling and disposition restrictions that untreated seed does not.
• Unit of measure handling that survives the whole chain. Bulk pounds at intake, units of a set kernel count for corn, bags for soybean, seed boxes and totes for delivery. Every conversion is a place where inventory silently drifts.
• Royalty accrual per variety licence at the correct trigger point, whether that is conditioning, bagging or sale, with a report the licensor will accept without argument.
• Carryover valuation, because seed that did not sell has a germination that will decline, a retest cost, and a class that may not survive the year. Treating carryover at full value is the most common accounting error in this business.

What it costs and how long it takes

Across the 2,000-plus projects Digital Heroes has delivered, this category prices as follows. A first release covering contracted production, harvest lot creation and full conditioning genealogy including blends and rebagging runs $65,000 to $140,000 and ships in 12 to 16 weeks. A full platform adding test scheduling and label age rules, certification agency reporting, treatment records, royalty accrual, grower settlement and carryover valuation runs $170,000 to $380,000 across 6 to 11 months.

What pushes it up: the number of crops, since a corn seed operation, a soybean operation and a small grains operation have genuinely different unit models and conditioning steps; the number of certifying agencies you report to, because each has its own forms and rules; licensed variety count and the complexity of royalty triggers; and plant equipment integration if you want conditioning yields read from scales and cleaners rather than recorded by hand. What holds it down: doing one crop and one plant first, and starting the build immediately after harvest so the genealogy model is exercised against a real conditioning season before the next one begins.

When you should not build

Do not build if you produce a small volume of public varieties, sell in bulk within one state, carry no licensed varieties and never blend across years. Your identity chain is short enough that a disciplined spreadsheet and a good tag file are honestly sufficient, and the capital is better spent on cleaning capacity.

Build when two or more of these are true. You carry licensed varieties and accrue royalties. You blend or rebag, which is where identity is actually lost. You produce across more than one certification class. You sell into more than one state and therefore live under more than one set of labelling rules. Or you have had a lot questioned, by a customer, a licensor or a certifying agency, and reconstructing its history took more than a day. That last one is the honest trigger, because the cost of a lost lot identity is the difference between seed price and grain price on everything descended from it.

How to choose a developer for a seed production build

Ask them to draw the lot genealogy on a whiteboard, then ask them to draw a blend of carryover with new production and a rebag. If the drawing survives, keep talking. If they reach for a parent child tree or a bill of materials, they have modelled discrete manufacturing and your first blend will break it.

Ask how class is derived on a blended lot. The correct answer is that the blend takes the lowest class of its inputs and the system enforces it, not that a user selects the class from a dropdown.

Ask how they will handle test expiry and label age rules that differ by destination state. This should be data with an effective rule per state, and the system should be warning you weeks before a lot becomes a problem.

Ask who owns the code, the repository and the cloud accounts, and settle it in writing before kickoff. At Digital Heroes the client owns the code from the first commit. Since there is no packaged fallback in this category, an inaccessible system is an inaccessible production history, and your production history is the value in your inventory.