Culling this autumn? ICBF already ranks your cows by future profit

ICBF ranks every cow in your herd by expected future profit, and the list is already sitting in your HerdPlus account. It is called C.O.W., short for Cow's Own Worth. It works for spring-calving, milk-recording herds only, costs nothing on top of HerdPlus, and updates itself every time a new record lands.

It is not new, whatever you read this week

ICBF put out a fresh piece on C.O.W. on 13 September. If you saw it described as a new tool, it is not.

ICBF was publishing research on C.O.W. while it was still at pilot stage in 2016. Teagasc was writing about farmers running it on their own herds in 2021. It appeared again in the 2023 Moorepark Open Day publication, written up by Margaret Kelleher of ICBF and Donagh Berry of Teagasc. It has been sitting in HerdPlus for years.

The reason you are hearing about it now is the calendar, not the software — late September is when spring-calving herds start drawing up a cull list.

That is the useful part. You are not being asked to sign up to anything or learn a new system. You are being told about a page you already pay for and have probably never opened.

Why does the milk recording sheet pick the wrong cows?

Most cull lists start the same way. You print the milk recording sheet, sort by yield, and look at the bottom.

Teagasc tested exactly that. George Ramsbottom looked at a 100-cow spring-calving herd in south Kilkenny with 71 mature cows in their third lactation or later. He pulled the bottom ten per cent by milk yield, then pulled the bottom ten per cent by C.O.W., and compared the two groups.

The cows with the lowest milk yield were predicted to produce 73kg of milk solids below the mature cow average. They were also younger, had higher EBI, calved earlier and ran lower cell counts.

Read that again. Sorting by yield handed him a group of young, fertile, high-EBI cows with clean udders. Those are cows you keep.

C.O.W. picked a different ten per cent — the ones genuinely likely to be less profitable for the rest of their time in the herd. "Milk yield alone is not a reliable measure to use when selecting cows," is how Ramsbottom put it in September 2021.

The point is not that yield is useless. It is that yield on its own cannot tell the difference between a young cow who is still filling out and an older one carrying a bad cell count and a late calving date. C.O.W. looks at both, and at everything else on her record.

What actually feeds the ranking

C.O.W. pulls live from the ICBF database and adds up three things for each cow:

1. Profit from her current lactation — milk volume, fat, protein, cull cow weight, milking speed, temperament, mastitis, lameness and somatic cell count.
2. Profit from her future lactations — what she is likely to earn if she stays, adjusted for her chance of surviving, getting back in calf, and the value of her offspring.
3. Net replacement cost — what it would cost to replace her, set against her carcass value.

Those three combine into one figure: what she is expected to be worth to you for the rest of her life in the herd. Every cow gets one, and they sort from best to worst.

Anything already flagged for culling, or scanned empty, drops to the bottom automatically.

It leans on the same genetic evaluation pipeline behind your EBI figures, then layers your own farm's records on top. If you have ever wondered what else is buried in the herd data ICBF holds on you, this is a decent place to start.

How is C.O.W. different from EBI?

This is the bit farmers get wrong, and ICBF states it plainly: EBI is for selecting animals for breeding. C.O.W. is for selecting females for culling.

The technical reason is worth 30 seconds. EBI uses half a cow's genetic merit, because a calf only inherits half from her. C.O.W. uses her full genetic merit, because it is judging her, not her progeny.

So a cow can sit mid-table on EBI and rank well on C.O.W., or the reverse. They are answering different questions. Keep using EBI to pick bulls — including through Sire Advice on HerdPlus — and use C.O.W. to decide who goes on the lorry.

What it costs

C.O.W. itself carries no extra charge. It is a profile page inside HerdPlus.

HerdPlus is not free. ICBF currently lists Dairy HerdPlus online at €100 a year plus 50c per cow calved, or €160 plus 50c per cow if you want it posted out. A 100-cow herd on the online option is looking at roughly €150 a year for the whole service, C.O.W. included.

Milk recording is a separate cost through your own provider, and rates vary by how many recordings you take. It is the one that decides whether C.O.W. is open to you at all: no milk recording, no C.O.W., and there is no way around that.

Where does C.O.W. fall down?

Three places, and they matter.

It only works for spring-calving, milk-recording herds. Autumn and split-calving herds are outside what it was built for. Suckler farmers get nothing from it at all.

It is only as good as what you have entered. It reads live from ICBF. If your scanning results are not in, the tool does not know which cows are empty, and empty cows are the single biggest thing it should be flagging. Teagasc is blunt about this: record the scanning results before you run it.

It does not know your yard. It cannot see the cow that puts you through the gate every morning, the one with a bad foot that has not been recorded, or the fact that you are short of grass and need to shift ten cows regardless of what any list says. It gives you a ranking. You still make the call.

Do this before you run it

A list built on stale records is worse than no list, because you will trust it.

1. Get your pregnancy diagnosis results entered. This is the one that changes the ranking most.
2. Enter any health events you have been meaning to — mastitis cases, lameness, anything treated.
3. Make sure your latest milk recording has landed.
4. Then open HerdPlus and pull the C.O.W. profile.
5. Walk the bottom fifteen or twenty with the list in your hand and your own eyes on the cow.

Teagasc's advice on timing is to run it in late autumn before drying off, or again before breeding starts. Running it pre-breeding stops you putting straws and labour into a cow you were going to cull anyway.

Frequently asked questions

How much is the gap between my best and worst cows actually worth? ICBF research put the difference in expected profit between the top 25 per cent and the bottom 25 per cent of cows on C.O.W. at €360 per cow per lactation, or €9,000 across a 100-cow herd. Treat that as the spread between your best and worst quarter, not money the tool hands you. It was validated on 2011–2012 national data, so the ranking logic holds better than the euro figure does in 2026.

Can I use C.O.W. on a suckler or autumn-calving herd? No. It was built and validated for spring-calving, milk-recording dairy herds, and ICBF has not extended it beyond that. If you are suckling, your EuroStar figures and your own records are still the job.

Does this replace a conversation with my advisor? No, and it is not meant to. C.O.W. gives you a ranked list. Whether you cull fifteen cows or five depends on your grass, your cashflow, your replacement rate and your winter feed position. That is a Teagasc advisor conversation, and the list makes it a shorter one.

The bottom line

If you are milk recording and spring calving, open the C.O.W. profile before you write this year's cull list — you are already paying for it, and the milk sheet on its own will hand you the wrong cows.

Enter your scanning results first. Then walk the bottom of the list yourself.

HerdPlus support is on 023 8820452 if the profile is not showing for your herd.