Three ways in, and what actually separates them
A trials unit, an advisory firm and a farm comparing two rates are not buying the same thing, even though one platform serves all three. This page says what each one gets that the others do not, and, at the bottom, what does not change whichever door you come through.
Replicated trials that survive review
Designs bound onto real geometry, a record with provenance on every reading, and analysis scoped to the design you declared.
What this one has that the others do not
- Randomised complete block, Latin square, split plot and factorial designs, placed with a seeded randomisation you can reproduce
- Analysis of variance with mean separation that refuses, out loud, when the design cannot support an honest error term
- Scales, grain gauges, moisture meters and NIR analysers read straight into the record, with barcode label sheets tying samples to the plot
If nobody will ever ask how the numbers were arrived at, this is more rigour than you need. Start with Farms.
Trials on client ground, across a whole book of clients
One organisation covering every member of staff and every device, and a hand-over the client's own agronomist can check.
What this one has that the others do not
- One organisation across staff and devices, so the person who laid the trial out and the person scoring it on Thursday see the same plots
- Shared field hardware that unlocks per person by PIN, with access revoked centrally when somebody leaves
- Machine execution on whatever the client happens to drive, so the trial does not require them to buy anything
Running trials only on ground you own, with one team? The extra organisation layer is overhead, and Research is the closer fit.
Prove what works on your own ground
Test the claim on your soil rather than somebody else's: a real comparison, driven by your own machinery, read back as an answer.
What this one has that the others do not
- Whole-farm mapping first, with boundaries imported or drawn and every note, pass and reading landing against the right field or part of one
- Strip comparisons rather than plot grids, sized to the machinery and the field you actually have
- RTK from any NTRIP caster, including the free public networks more than twenty US states operate, with no correction-source lock-in
Publishing the result, or defending it to a third party, asks for the design and analysis on the Research page.
One platform underneath all three
These are three ways in, not three products. Whichever you start from, the same platform is underneath, and moving between them is a change of emphasis, not a migration.
Offline is the normal case
Cached maps, local projects, and recording that never needs signal. Sync catches up when you are back in coverage.
Export is never gated
Every format, every plan, whether or not a subscription is active. Stop paying and the app goes read-only while export keeps working.
Your own machinery
ISOXML to the terminal in the cab, or Fieldbase firing the outputs itself. No brand requirement in either direction.
Set up your next trial in Fieldbase
Free to try, with no card required. The drive simulator lets you test a complete field before you connect a GPS receiver.