Be a Beauty Queen Instead of a Scream Queen

Profile Picture of Post's Author Erin Mosley | Posted May 6, 2024

To me, data can be a beautiful thing. Yes, I am talking about all those numbers that are generated every minute, or even second, that tell the story of a dairy farm. When not maintained though, data can be more like a horror movie. You know the ones where only the monster makes it out alive. I know data isn’t always a straight line, but I also know that constant zig zags in dairy data can be frightening.

Everyone these days is talking about data. Cow data, feeding data, milking data, rumination data… If you visit a dairy today, you will likely find multiple specialized computer programs used to control and record data from multiple systems throughout the day. There are so many data collecting programs on dairies, it can be overwhelming to think about and manage all the data being collected, much less know where to start when looking at it.

Many times, these programs are feared and treated using a hands-off approach. You know, if it ain’t broke don’t fix it, or even click any buttons. We often forget that a computer is still just a machine. These programs really do need some routine maintenance. What do I mean by that? Well, pretty much the way it sounds. Someone must look at the data being collected. Some companies are working to create data dashboards to help pull all this data together, but I am focused on the source of the data. If the data being collected is not correct, your dashboard data – no matter who makes it – won’t be correct either.

Imagine this scenario – you have a pen of cows that have been slipping in dry matter intake. You get concerned and call up your nutritionist to see what they have to say. Ultimately, you decide to check out the feeding management program. That’s where I come in. My job is to look at this data, no matter the program, no matter the dairy. Usually, I am looking at data from an observational standpoint, just figuring out what the patterns are on the dairy over a period of time. But today, I get the call to see what I can find in the feeding program to identify why this specific pen is down in intake.

Ok, so now I’m an investigator. Digging through daily loads, daily pen feeding time, looking at loading and delivery accuracies, evaluating recent ration changes, checking forage DM values, evaluating pen count changes, pen target DM changes, pen refusal values, intakes, inventories… When the loads are clean, and the data is consistent and correct, identifying potential issues is much clearer. BUT when drop times and intakes look like constant zig zag lines day to day or refusals or other data is inconsistent or missing, determining possible issues becomes much harder.

So, what do you do when you find the monster called questionable data? Beautify it! Start with one issue at a time. Identify why first feeding times are variable, stick to a schedule for routine forage DM entries, establish guidelines for daily pen target DM intake changes, routinely evaluate feeder loading and delivery accuracy and communicate goals to feeders. Then, keep looking at the data. Set a time each day or each week to identify issues as they come up, and not months after the fact. Pretty soon you will be a pro at your program. You will slay the monster. Your data will be beautiful, and you can trust it – because you know that you have verified it along the way.

You can find a PDF of this blog post here.

    Categories
    Latest Articles
    Archive