With all of the data available to today’s dairy producers, the value of that data is easily lost in the shuffle. Data is raw material that should be used to “drive understanding,” Robert Goodling Jr. said in a Penn State Dairy Nutrition Workshop session.

Freelance Writer
Tamara Scully, a freelance writer based in northwestern New Jersey, specializes in agricultural a...

“We need to think about, ‘How are we going to use [data]?’ If we don’t go any further, what has that really gained?”

The workshop, “The Value of Data Monitoring to Examine ‘Employee’ Performance,” was a collaborative effort between Goodling and Kenneth Griswold, senior technical service provider – dairy, Kemin Industries.

With all of today’s technology, data is readily available. Too many producers adapt the technology – but don’t utilize the data. But that data can offer valuable insight into cow performance.

It can help to pinpoint management changes that can be made to help a dairy meet cash-flow goals and capture lost milk money.

Advertisement

Producers aren’t going to benefit from looking at data without any purpose in mind. Data is used to answer a specific question. By “seeking the right questions” and finding the correct datasets to provide answers, dairy farmers can take advantage of the numbers and learn to examine them to fulfill a given need.

Data types

Knowing what data is needed to help answer your questions – and where to find that data – is the first step. Dairy producers can collect data from DHI records, bulk tanks and milking parlors.

Knowing the limits of each dataset and the information each offers is key to monitoring performance. Milk parlor data can provide a lot of individual cow-level information in real time, but available data varies from producer to producer and might be limited to milk yield.

Bulk tank or co-op data drives income, as it is directly related to the milk check, but it is limited to herd-level performance. DHI records can provide cow- or group-level data – yet is a time-limited, periodic snapshot of herd performance.

Bulk tank herd-level data can reflect changes in feed quantity and quality, environmental stressors and herd health. This data can be available at varying frequencies – daily, weekly or monthly.

Parlor data is good for monitoring individual cows. Some farms may only have milk weights available, while others will have not only yield data but composition and cow behavior measurements as well.

DHI data is normally collected monthly and offers a snapshot on the group level and the cow level. There are more than 100 metrics available, often leaving producers wondering how to process all of the information and make it usable.

“Understanding what each data source provides and how the three integrate to provide an overall herd performance picture” is essential, Goodling said. “Using just one source of data can limit the ability to effectively monitor or investigate a herd’s performance. Utilizing all three will provide a more complete picture of herd performance.”

Data selection

Making data-based changes can positively impact a farm’s breakeven point. Data can be used to examine where management changes can improve performance, but the correct and most complete data must be utilized.

For example, the amount of milk per cow per day has many component parts. Forages, cropping, feed management, ration formulation and cow comfort are some of those. Changes in any of these will impact production totals.

When examining available data, it’s important to understand the time frame represented by the numbers. If changes were made on the farm, data examined too close to that change may not reflect its impact. If data is from a month ago, it won’t reflect changes made on the farm last week.

Knowing when data was collected and how long any changes that occurred will take to impact results are both needed to utilize data correctly. Whether the changes impact the entire milking herd or just one cow or pen will help direct what data needs to be examined.

If a forage change is occurring on the farm, what data should be examined? DHI data won’t be too useful here due to the lag time. What farmers want to know is the immediate bulk tank response to the feed change, Goodling said.

But not all groups of cows will provide helpful data. The data from cows whose milk production will be affected by the change right now is needed.

First-calf heifers will respond with growth but not changes in milk. Second-lactation cows greater than 100 days will show minimal response to feed changes.

Cows less than 30 days postfresh will be off on their DMI and not provide useful data. But cows from 31 to 100 days postfresh will provide a usable response to the feed changes and show how milk production was impacted.

The data from that specific group of cows can answer the question about the response to the forage change. Focusing on parlor data from this group of cows will show immediate changes in response to the changes in feed management that occurred.

“We’re just going to get lost in the system,” if all the generated data is examined without context, Goodling said. “Some of the stuff that you can generate may not give you any help” with the issue at hand.

All data can be inaccurate. If machines are not properly calibrated, the data will not be correct. Snapshot data can show things that already happened, such as a temperature change, which have already been rectified and are no longer pertinent. Data that reflects averages can overlook outliers.

“Averages can mask a lot of things,” Goodling said.

With “endless possibilities” in data availability, it can be a big challenge to find and examine the data that can assist dairy producers in making the optimal management decisions for their farms. Knowing what data is needed to answer specific questions and where to go to find it is key.

Learning to utilize relevant data from the various available sources can provide an integrated, informed look at cow performance, improve management and enhance the bottom line.  end mark

Tamara Scully is a freelance writer from Columbia, New Jersey