Two flocks can average the same weight and make very different money. One is tight — most birds within a pound of the average, catching evenly, grading out clean. The other is wide: big birds and small birds in the same house, feed wasted on the tail end, and a settlement that shows it. Uniformity is what separates them, and coefficient of variation (CV%) is how you measure it without guessing. This guide gives you breeder CV% targets by age, a sampling protocol that produces a number you can trust, and a decision tree for when the flock is wider than it should be.
What CV% Actually Measures
CV% — the coefficient of variation — is the standard deviation of your sample weights divided by the average weight, expressed as a percentage. It describes how spread the flock is relative to its average, which is what makes it comparable across ages and flocks:
- CV% = (standard deviation ÷ average weight) × 100.
- Low CV% = tight flock. A CV of 8% on a 1,500 g average means most birds fall between roughly 1,300 g and 1,700 g.
- High CV% = wide flock. A CV of 15% on the same average means a substantial tail of birds far below — and far above — the target, and the birds below are eating your margin.
- CV% is scale-free. That is why an 8% CV at day 14 and an 8% CV at day 35 describe the same quality of flock management, even though the weights are completely different.
Uniformity percentage (% of birds within ±10% of the average weight) is the older, simpler measure you will also see. CV% carries more information — it describes the whole distribution, not just one band — so use CV% as your working number and translate to uniformity percentage when a buyer or integrator asks for it.
Breeder CV% Targets by Age
Source: Cobb-Vantress Cobb500 Broiler Management Guide and Aviagen Ross 308 Management Handbook uniformity expectations. These are breeder-published targets under standard management — field results commonly run wider.
| Age | CV% Target | Uniformity (±10%) | What a Wide Flock at This Age Usually Means |
|---|---|---|---|
| Day 7 | ≤ 8% | ≥ 80% | Brooding temperature spread, feeder/drinker access for small chicks, chick quality from the hatchery. |
| Day 14 | ≤ 9% | ≥ 80% | Feed and water access still uneven — pan height, nipple pressure, litter caking under nipples. |
| Day 21 | ≤ 10% | ≥ 85% | Established pattern: early access problems now showing as size gaps. Investigate before week 4. |
| Day 28 | ≤ 11% | ≥ 85% | Subclinical disease, feed phase transition stress, or ventilation stratification (cold spots). |
| Day 35+ | ≤ 12–15% | ≥ 85% | Some widening is normal as growth slows; more than 15% usually means an unresolved problem from earlier weeks. |
- Cobb-Vantress. Cobb500 Broiler Management Guide — uniformity expectations by age. cobb-vantress.com
- Aviagen. Ross 308 Management Handbook and Ross 308 Broiler Performance Objectives (2022). aviagen.com
How to Sample for Uniformity
The protocol matches the growth chart sampling guide — the same sample produces your average weight, your CV%, and your comparison to target:
- Sample 30–50 birds per house, randomly selected. More is better for CV% — a 50-bird sample describes the distribution far better than 20. Do not pick visibly sick or cull birds; do not pick the birds nearest the door.
- Weigh individually, not as a group. A group weight gives you an average with no distribution — no CV% is possible from it. Use a hanging gram scale or certified digital platform.
- Weigh at the same time each sampling day — mid-morning is ideal, after crops have emptied overnight and before afternoon feed fills them. Time-of-day shifts in crop fill inflate apparent variation.
- Calculate three numbers: average weight (g), standard deviation (g), and CV% = (standard deviation ÷ average) × 100. A spreadsheet or phone calculator handles it; the calculation is one division.
- Compare against the table above for the flock's age, then against your own previous flocks at the same age. Your own history is the most honest benchmark you have.
- Log all of it: date, house, number of birds sampled, average weight, standard deviation, CV%. A CV% with no record is an anecdote; a logged CV% series is how you prove a pattern — or prove there isn't one.
Sampling weekly from day 14 onward catches the window where intervention still works. A CV% that jumped 3 points between two weekly samples is more useful than a single "good" or "bad" reading — the trend tells you when the problem started, and the earlier that is, the more options you have.
When CV% Is Too High: Decision Tree
Before diagnosing the flock, check the measurement. Was the sample random or convenience-based? Were crops full for some birds and empty for others? Is the scale calibrated? A flawed sample produces a wide CV% that does not exist — and sends you chasing a phantom problem while a real one compounds.
Feed pan height and fill level for bird age, feed flow to every pan (caked feed, bridging in the line), nipple height and pressure by age, and drinker line level. In nearly every wide flock I have seen diagnosed properly, the small birds were the birds with worse access — not worse genetics. Walk the house at bird level and look at what the bottom third can actually reach.
Brooding temperature spread (cold corners, drafty pads, heater cycling), ventilation stratification, and litter condition. A house with cold spots grows two flocks — one in the warm zone, one outside it. Temperature sensors at bird level across the house, not just at the controller, reveal this.
Subclinical disease widens flocks before it kills birds. Respiratory signs, litter quality, coccidiosis history, and any treatments given — a treated and an untreated subpopulation in the same house is a CV% problem with a medical cause. Call your veterinarian with the CV% series and mortality history attached.
Use the flock's average weight and CV% against the growth chart to project the sale weight of the bottom quartile. If that quartile will not reach sale weight on the current schedule, the decision is whether to extend grow-out (costing feed per the feed chart) or to accept the shortfall and re-forecast settlement. Run the numbers on paper — the answer depends on feed price, bird price, and days remaining, not on a rule of thumb.
Uniformity and Money: Reading the Settlement Impact
Uniformity shows up in the numbers that matter:
- Feed wasted on the tail: A wide flock's small birds keep eating without reaching sale weight. At $0.50/lb feed, every extra week of those birds is real money for birds that may never pay it back — the FCR guide shows how much the flock-wide conversion suffers.
- Uneven catch and processing: Birds far above target grade out worse; birds far below cost you on head count. Integrator settlement sheets reflect both.
- Cost per pound produced: The cost per bird and cost per pound metrics on the profitability dashboard move with uniformity more than most growers expect — because the wide flock's true cost is spread across fewer saleable pounds.
- What good looks like: A tight flock at slightly lower average weight can out-earn a wide flock at higher average weight. That is not a theory — it is what the settlement comparison shows, flock after flock, once the numbers are logged consistently.
Tracking Uniformity in Poultry Log
When sample weights are logged per house in Poultry Log, each sampling event stores the individual weights — so average weight, CV%, and the comparison to breed target are calculated for you, per house, per date. The CV% series across the flock's life is the part spreadsheets rarely survive: it shows when the spread started widening, which house it happened in, and whether the last intervention worked.
Paired with feed, water, mortality, and treatment logs on the same timeline, a CV% trend becomes readable against its likely cause — and every flock's uniformity history feeds the profitability dashboard, where you can finally see what wide flocks have been costing you across seasons.
- 2026-10-05: Initial version. Breeder CV% targets by age, sampling protocol, decision tree, and settlement impact framing.