Every livestock record is a choice about resolution: does this entry describe one animal, or does it describe a pen of forty that are being treated as one unit? Neither choice is wrong in itself. A group record kept where an individual one was needed hides exactly the animal you needed to see. An individual record kept for every event on a two-hundred-head commercial herd is a workload nobody sustains, and a workload nobody sustains gets abandoned, which is worse than either.
The right answer is not "always individual" or "always group." It is matching the resolution of the record to the resolution of the decision it will eventually support.
This decision is easy to get wrong in both directions, and both mistakes look reasonable at the time. Recording individually out of caution feels thorough right up until it becomes a second full-time job nobody has hours for, at which point the individual records quietly stop being kept at all. Recording at the group level for speed feels efficient right up until one animal in that group needs to be told apart from the rest, and the record has no way to do it. Neither failure announces itself in advance; both show up only when a specific question finally arrives.
A group is the right unit when the animals are actually interchangeable
A pen of steers bought together, fed the same ration, moved through the same gate on the same day, is a legitimate group for most purposes — a feed allocation, a pasture move, a vaccination protocol applied to the whole pen at once. Recording these events at the group level is not a shortcut that loses information; it is the accurate description of what happened, because nothing distinguished one animal's experience from another's. Forcing an individual entry for each of forty animals that all received the identical thing on the identical day adds workload without adding truth.
The test worth applying before defaulting to a group record: would every animal in this group give the same answer to a question I might ask later? If yes, group-level is correct. If the honest answer is "probably, but I'm not sure," that uncertainty is itself the warning sign that the group is hiding variation.
A group record hides the one animal you actually needed
The failure shows up when the group is not really uniform, or stops being uniform partway through. A pen where one animal gets pulled for a treatment the rest do not receive, tracked only at the group level, now contains a record that is true for thirty-nine animals and false for one — and there is no way to tell which one from the record itself. A cull or mortality event inside a group is the sharpest version of this: the group's aggregate weaning weight or gain figure now includes, or should exclude, an animal that left partway through, and a purely group-level record has no field for saying which one or when.
This is also where a withdrawal obligation becomes dangerous at the group level. If one animal in a pen was treated and the rest were not, and the record only ever speaks of "the group," a sale decision made against that group record either withholds every animal unnecessarily or — worse — sells the one that should not have moved, because nothing in the record distinguishes it from its pen-mates. This is the same discipline covered in withdrawal and residue records, applied to the question of which unit a treatment record is written against.
The two levels are not a single choice made once
A working herd rarely picks one level and uses it for everything. The more common, and more defensible, pattern is individual identity kept for the events that matter per-animal — birth, breeding, weaning weight, treatment — with day-to-day handling, feeding, and pasture moves recorded at the group level, because those really are uniform across the pen. What makes this coherent rather than two disconnected systems is a group identifier attached to every individual record, so an individual animal's history can always be read in the context of which group it was part of on a given date, and a group event can always be exploded back out to the individuals inside it if a question demands that resolution later.
What breaks this pattern is inconsistency about which events get which treatment — recording some treatments individually and others, for no principled reason, at the group level, so that six months later nobody can predict which resolution a given record was kept at without opening it and checking.
Group composition itself needs a date
A subtler failure: groups are not static. Animals move between pens, get sorted by weight or sex, join and leave a group throughout a season. A group record that does not carry a start and end date for each animal's membership cannot answer, months later, which specific animals were actually in "the north pasture group" on the day a feed allocation or a treatment was logged against it. Without that date, a group-level record is only as good as everyone's memory of who was where, which is the exact failure a group record was supposed to avoid.
The fix is unglamorous: a start date, and an end date when one applies, on every group membership, recorded at the moment an animal is sorted in or out. It is a small habit compared to the payoff — it is what turns "the north pasture group" from a name everyone half-remembers into a precise, reconstructable set of animals on any given day, which is the only version of a group that a treatment or a feed record can safely point back to later.
Choosing the resolution in practice
The practical rule: default to individual identity for anything that could plausibly need to be traced back to one animal — a treatment, a sale, a breeding event — and use group-level entries for genuinely uniform, whole-pen activities, always with a group ID attached to the individual records so the two stay reconcilable. Farm40 lets every individual animal record carry a current and historical group assignment, and its read-only assistant can be asked which animals were in a given group on a given date — up to 10 questions a month at no cost. Its limit is the one that applies throughout this page: it can only report the group assignments that were actually entered. A move that happened but was never logged is invisible to it, exactly as it would be to a spreadsheet.
Choosing the right resolution is a foundational decision inside the livestock management guide, because every other record type on this list — weight, breeding, treatment, feed — inherits whichever resolution the group decision set.