SIOUX FALLS, S.D. – Adapting and using technology appropriately for each cow is critical when gathering data and acting on it for the betterment of a dairy.
    Industry experts shared guidance based on their experiences with new technology and data with dairy farmers during the Data and Innovation Summit June 30 in Sioux Falls.
    “As an industry, we have the dilemma of two ways to think about decision making,” Dr. Jason Osterstock said. “There’s intuition which is steeped in tradition and heritage, and then there’s data-driven which we practice a lot in milk recording, herd recording.”
    Osterstock, with Zoetis, presented “Unlocking the power of data for the hardest working cows in the business” during the summit.

The power of cross-referencing data
    Since farms are managed with a whole-herd approach, Osterstock gave examples of how technology and data derived from such technology help farmers make more informed decisions for each individual cow. At the same time, the whole herd can be kept in mind when considering the long-term goals of the farm.
    “At the time of my training in preventative medicine herd health, the individual cow wasn’t a fad but it was, fundamentally, the approach,” said the veterinarian. “As herd sizes expanded, we didn’t have the luxury of practicing individual medicine when thinking about managing populations. Twenty years later, we’re driving back to individual cows with technology.”
    He cautioned the use of data from specific technologies for only the obvious purposes.
    “It’s imperative we break down the silos in our data,” Osterstock said. “Don’t think of genetic technologies for genetic data, parlor for parlor. We need to think about how this data comes together for a data ecosystem.”
    For example, Osterstock spoke of using genomic information for productivity but also tracking incidence of disease survival in a herd. There is opportunity to make appropriate decisions knowing the accuracy of genomic predictions, particularly when comparing high and low genomic animals.
    Osterstock’s presentation showed stark differences between high and low genomic animals regarding mastitis and displaced abomasum.
    “Those predictions are almost scary at how well they work,” Osterstock said. “If you start to see a DA, from a genetic perspective, you should not be getting DAs; that’s the yellow canary that says there’s a problem there.”
    Using such information in vast ways allows farmers to proactively manage their herds while considering both genetic and environmental risk factors.
    
Know how to use the data
    Having a plethora of information can lead a dairy farm in the right direction, but using that information poorly can be detrimental to the business.
    “Unfortunately, if data is used incorrectly, bad things happen,” Dr. Michael Overton said. “Maybe you’re looking at the wrong data and intervening with a problem that doesn’t exist or you fail to intervene with a problem that could be fixed.”
    Overton, with Zoetis, explained how to appropriately use data in his summit presentation, “Do’s and don’ts of interpreting farm data: Avoiding common data mistakes.”  
    Before farmers are ready to evaluate the data, they should first formulate a question. This will allow data to be collected and trends to be reviewed before developing an action plan with farm consultants such as the nutritionist or veterinarian.
    Overton warned of variations in the data sets.
    “It is the bane of my existence,” he said. “It creates clouding when interpreting what’s happening in the herd’s performance. We have to make sure we can ascertain what we measure represents what the population is doing.”
    Dairy farmers should also be cautious of bias in their data. For example, a selection bias was presented when asking whether older cows make more milk.
    Overton told of a case of 15,000 cows followed over five years. The surviving cows gave more milk each lactation and it increased with the average population.
    “We weren’t comparing apples to apples because cows dropped off with each lactation,” Overton said. “Yes, older cows made more milk, but what we observed is selection bias and keeping better animals around longer.”
    Averages also do not tell the whole story of a herd.
    Overton gave the example of Michael Jordan walking into a bar during happy hour.
    “Every person instantly becomes a millionaire, on average,” he said. “In reality, the average net worth could be $85,000, but one outlier changed the perspective of what you would interpret the average to be.”
    Considering some of the ways data can be misinterpreted, Overton also stressed the importance of reviewing industry benchmarks and other farms’ data points. He said milk production and reproduction can easily be benchmarked with values consistent across the industry, but other values vary greatly between facilities.
    The same goes for treatments.
    “The reality is that each herd had a reason to make those decisions,” Overton said. “Looking at the historical references on the farm will give a conscious insight to the right decision.”
    With available technology in the industry, the data holds a wealth of potential for dairy farmers to further their success.
    “Farm data is a collection of bits of information used to improve the herd and farm,” Overton said. “By itself, it’s largely useless. Look at it all before taking corrective action.”
    Osterstock agreed.
    “Be vigilant and optimistic in what you can do,” he said. “Exert influence, be creative; there’s a lot of great stuff coming that will fundamentally change the way we dairy.”