In the 1990s, a new feed and nutrition customer of mine had two printed reports ready for me on my regularly scheduled visit. The first was a printed ration that included a feed cost per cow. The second was a cash flow or some other financial report from the accountant that included a feed cost per cow and per hundredweight (cwt.). Her question: “Why doesn’t the price per cow on this ration agree with what the accountant’s reporting shows as actual feed cost?”
Having recently transitioned from my beef cattle upbringing to my new job as a dairy nutritionist, I was probably caught flat-footed with this question. In my recent but very similar conversation, this now well-understood list of potential disconnects between formulated feed cost and actual hasn’t totally solved this issue.
We have gotten better at this. Instead of printed papers with a dollars per cow per day indicated, we use a dynamic model that is reflective of nearly every factor that can influence the actual cost of feeding animals on the dairy.
I will always first be a cost control manager when it comes to building rations, motivated to find ways to supply the same nutrients but at a lower cost. This effort is to find the simple economic sweet spot to maximize margin per cow, per stall, or per whatever denominator you choose. Variable production with several levers like 3x or 2x, overcrowding, and genetic investments, obviously are critical in finding the maximum margin, but with whatever production approach you choose,
To sort out these potential areas of feed cost disagreement, we should start by separating forages from grains, by-products, minerals, and other similar ingredients. Forages usually have the higher shrink risk. Not that grains don’t blow away in the wind, but forages, and silages in particular, experience higher shrink. The other area of uncertainty in silages is the true dry matter percent of a pile. Here, there are many potential areas of miscalculation, and the truth is probably not known until the financial statements show up several months later.
Non-forage ingredients have shrink risk as well, but it is less often added to the cost per ton used in ration software. The values could range from 5% to 10% and make a difference in predicting true feed cost. Using a more detailed shrink approach put forth in my article in the January 10, 2025, issue of Hoard’s Dairyman, titled, “Calculating shrink isn’t just for silage,” that I called “differential shrink,” we can use the formulation software to help predict the actual impact of variable shrink risk on a per ingredient basis. The point here is that high-cost ingredients that also happen to have a higher shrink risk and a higher feed rate will have a more significant impact on potentially underestimating actual feed cost.
Now for an accounting risk. The values that often get compared are cost per cwt. on a financial statement. Though it’s an old dairy economics term that hasn’t really aged well, it is still common to see in a dairy’s financial assessment. When considering this number compared to a cost that might show on a report from my ration software, it is critical that we use the same pounds of milk factor for the denominator in this simple equation. In most ration model reports, the denominator used will be what the ration is formulated to support — not actual milk flow. There are many legitimate reasons why these two predicted values and actual milk may not agree. Most accounting feed cost calculations will include the hospital and any other off-line cows in the math, lowering the milk per cow. And, when calculating feed cost per cwt. including dry cows, are close-up heifers included? There is no standard for this question in dairy accounting.
It is no small task to be sure that any nutrition formulation, on-farm feeding software, or feed economic tracking models have the correct price per ton listed. Confusions here could be from having multiple contracts for a particular ingredient that could either be entered as a weighted average or feed the oldest contract first. Additionally, spot loads might be required if contracted loads are either fed up or perhaps not available.
If you have a feed cost per cow printed on a piece of paper in 1995 or a PDF in 2025, what level of intake is used? This is the most important factor that we account for in a dynamic economic model we use, and one that I hope we get right most of the time. If a static printed ration report indicates a cost of $7 per cow, that number will be right only for the intake used in that ration. If that $7 is at 55 pounds of dry matter intake (DMI), and actual intake is 58 pounds, then the actual feed cost is $7.38 per cow. Thus, we must report feed cost using actual updated intakes, and if there are multiple rations on one farm, each with a different feed cost, we must accurately track how many cows are assigned to what ration and how much did they actually eat. Another potential area of underestimation is which group of animals pay for refusals. And, what about shrink that is not at the feed center but on the feed lanes from blowing snow, dust, and excessive rain? All of this creates risk to under predict feed cost.
The path to improvement is through communication. Perhaps, just like we learn over time how to best estimate shrink on silages by what actually happened at a particular dairy, we could create a feed cost mover that makes educated guesses on actual versus formulated feed cost. If I was a banker and in the business of lending money to a dairy for a new start-up, I would have such a number in mind. I expect this is already in place. What may be missing is nutritionists having this same factor in mind when formulating and reporting a formulated feed cost for each new ration formulated.
