Data, straights, yield and fertility

Robots produce a huge amount of data. Visits, yields, refusals, rumination, cake intake, cow by cow, every day. Most farms use a fraction of it.

This farm was also feeding a blend. Easy to handle, but you're paying someone else to decide what goes in, and changing the ration means waiting on the next load.

The data is one of the biggest advantages a robot herd has over a parlour. It shows you which cows are under-visiting, which are overfed, and which are slipping before you'd ever spot it by eye. Left unused, it's just numbers on a screen.

Blends have their place. But the recipe is fixed, you can't always see what's in it, and you can't change it quickly when forage or prices move. For a farm with the storage, straights give far more control.

First, the data. We went through the robot figures properly to find where cows were being held back, breaking the herd down by stage of lactation and by lactation number. Numbers are only useful if they change something, so every review ended with a decision.

Then we moved off blends and onto straights. A premix is made up weekly, so raw materials don't need mixing every day, and it can be altered at any time to match performance. We know exactly what each cow is getting and what it costs. It has also let the farm start growing more of its own crops, which helps with rotations and weed burdens.

It also means the ration can change quickly. If a new clamp is opened, forage analysis shifts or a feed price moves, we adjust straight away. Buying straights also opens up forward buying when prices are right.

Feeding a much cheaper premix down the feed trough has allowed us to feed a lower rate in the robot without compromising performance. The farm also has two feed bins connected to the robots, so we can feed two types of cake depending on stage of lactation and performance.

Fertility was the other big focus. Days in milk is always a challenge in a robot herd. Cows milk well for longer, and it's easy to let them run on. Getting energy and protein right in early lactation, and keeping the rumen stable, gives cows the best chance of getting back in calf.

Over three years, yields are up 20%, with better control over the ration and more control over cost.

Fertility has improved alongside it. More on that below.

Three years ago this herd had too many cows running on, too many services per pregnancy and too many three-week periods where hardly anything got in calf. Here's how each year compares, from October to September.

Pregnancy rate has nearly doubled. Conception rate is up by half.

The first year cleared the backlog. More cows were served, and the number of open cows waiting to be served fell from 105 to 76.

The second year is where the cows started holding. Conception rate jumped by a third in a single year. The farm used a third fewer services than the year before and got almost the same number of pregnancies per three-week period.

More pregnancies from fewer straws. In 2025–26 the herd got 180 pregnancies from 412 services. In 2023–24 it took 582 services to get 170. That's 170 fewer inseminations for 10 more pregnancies, saving semen, labour and time.

Open cows are down by a third. They've fallen from 46% of the herd to 30%. Fewer cows running on means more of the herd in the profitable part of the lactation.

It's consistent. Three years ago, most three-week periods were poor. Now it's the exception: 3 periods out of 16, down from 10 out of 17.