Five minutes at seven o'clock

Camping table, laptop, coffee, in the middle of the farm. I’ve been out here since six. At seven a cattle truck comes in with two trailers. Two feeders start on the ration. Two tractors are already out. Three buggies head down — two to muster, one for the bunks. John goes past me on the motorbike, lifts a hand, and follows them down. All of that inside five minutes.

THIS WEEK

Yesterday the same yard was empty. It was Sunday, and when I worked out here Sunday was a work day. Marq spotted my car before I was out of it, surprised to see me — Andrew had told John I was coming, and the message stopped there.

Before that, Brisbane. What I brought out here is a working version of LPOS — the first one I would put in front of someone. Tamir and I rebuilt it in two weeks and moved it onto our own servers; the first attempt took me three months. On the way west I stopped in Dalby, in a shop where I used to weld cattle crates, and sat in the owner’s garden while he described the software he wishes he had.

WHAT I SAW

Dinner with John and Marq ran long. I showed them what I have built.

They got excited — and not about the part I was proud of. They got excited about the basics. Records in one place. Entries you don’t type twice. Numbers that know about each other. The things any software can do.

None of that was news to me. Doing the simple things properly is the floor, not the ceiling, and I had assumed anyone selling into these yards had the floor covered. So I built to win on the hard part — the reasoning, the forecasting — and never got to show it.

Then they told me about someone who came out to pitch them something. Thirty minutes in, they could tell he had never stood in a yard at six in the morning, and they stopped listening. What they said they liked about mine wasn’t the software. It was that somebody asked what breaks their day and then went off to fix it.

Meandarra, Monday, just after six.

WHAT IT CHANGED

I had the order right and the competition wrong. If the basics were already handled out here, they would be the price of entry and nothing to talk about. They are not handled. Which makes them the offer: records in one place, no double entry, everything linked, and as few clicks as the job allows, so it fits the way the work already runs.

The seven o’clock version — the factors weighed, the scenarios run, a recommendation with the alternatives still under it — is still where this goes. It just isn’t what wins the first conversation.

STILL OPEN

Next is another farm, the same screens, the same questions: do they have the same problems, or have I just learned one family’s habits? I would rather find that out than build another feature on an assumption. In the end it has to suit them, not me.

Talk soon,

Yannik