Production scheduling for CPG means sequencing high-mix, short-run products across shared lines under four pressures at once: sequence-dependent changeovers (including allergens), shelf life and code dates, promotion-driven demand spikes, and retailer on-time-in-full targets. The sequence, not just the quantity, decides whether the week works.
Scheduling a machine shop is mostly about capacity and due dates. Scheduling a consumer packaged goods plant adds constraints that machine shops never see: a run order that must respect allergen rules, product that expires, demand that doubles when a retailer runs a promotion, and fill-rate penalties measured by the case. This guide covers what makes CPG scheduling different, how to structure the week, and where software and AI actually help. It pairs with our general production scheduling primer and the wider view of CPG manufacturing operations.
What makes CPG production scheduling different?
CPG scheduling is line scheduling under sequence rules, not job-shop routing. Most CPG plants run a handful of high-speed lines, each capable of many SKUs, so the core decisions are which SKUs run on which line this week, in what order, and in what batch sizes. Capacity math matters, but the sequence within each line often matters more, because changeover time in CPG depends heavily on what ran before: a dark flavor after a light one is a rinse, the reverse is a teardown, and an allergen in the wrong slot forces a full sanitation cycle. Add shelf life, which punishes overproduction, and promotions, which punish underproduction, and the scheduler is squeezed from every side.
The practical consequence: CPG schedules are built around run strategy, deliberate, repeating patterns of what runs when, rather than rebuilt from scratch each week. The plants that struggle are usually the ones treating every week as a blank sheet.
How do changeovers and allergens drive the sequence?
In CPG the changeover cost between two products depends on the pair, so the run order on each line is chosen to make transitions cheap and safe: light to dark on flavors and colors, allergen-free before allergen-containing, with full sanitation breaks only where the sequence forces them. Running products in the wrong order can double the week's changeover hours without adding a single case of output. This is the discipline covered in depth in changeover sequencing, and it has a food-safety edge: allergen changeovers are not just slow, they are regulated. Scheduling all sesame-containing SKUs at the end of the day, followed by a validated cleaning, is a scheduling decision that doubles as an allergen management control.
Two practical rules follow. First, sequence rules belong in writing, a transition matrix that says what may follow what and at what cleaning level, so the schedule does not depend on one planner's memory. Second, shrinking the changeovers themselves through SMED buys scheduling freedom: every hour cut from a teardown is an hour the scheduler can spend on shorter, fresher runs.
How do shelf life and code dates constrain the schedule?
Shelf life caps batch sizes and forces frequency. A product with a 45-day code and a retailer that demands 75% of shelf life remaining at delivery gives you roughly ten days from filler to truck, which means big, efficient campaign runs can quietly produce cases that are unsellable by the time they ship. The scheduler is really allocating freshness: run lengths must clear demand without outrunning code dates, and inventory must rotate first-expiry-first-out. This is why CPG leans make-to-stock but cannot simply maximize run length the way a durable-goods plant can; the trade-offs are covered in make-to-stock vs make-to-order and batch production.
The failure modes are symmetrical. Runs too long create aged stock, markdowns, and waste. Runs too short burn the week in changeovers and risk shorting orders. The balance point moves with demand, which is why the schedule has to be rebuilt against a live demand picture, not last quarter's averages, the connection to demand planning is direct.
What is a product wheel and why do CPG plants use one?
A product wheel is a fixed, repeating sequence for a line, every SKU gets a slot in a cycle, in changeover-optimal order, and only the run lengths change with demand. It converts the hardest scheduling problem (what order to run) into a solved one, so weekly scheduling becomes sizing the slots rather than re-deriving the sequence. High-frequency SKUs appear every cycle; slow movers appear every second or fourth turn of the wheel. The wheel also levels the load in the spirit of heijunka, and long-run variants shade into campaign scheduling where changeovers are brutal. Promotions are handled by growing a slot, or pre-building ahead of the event within shelf-life limits, rather than by tearing up the sequence.
How do retailer OTIF targets shape the schedule?
On-time-in-full programs turn the schedule into a service commitment: major retailers score suppliers on whether each order arrives complete and inside its delivery window, and shortfalls carry chargebacks. For the scheduler this changes the objective. A plan that maximizes line efficiency but shorts two orders costs more than a plan with an extra changeover that ships everything, so due-date protection outranks changeover savings whenever the two collide. In practice that means promised orders are pinned to schedule slots with slack ahead of their ship dates, and the week's flex, overtime, a weekend shift, a slot borrowed from a make-to-stock run, is spent defending committed volume first.
OTIF pressure is also the strongest argument for scheduling from live data. A line running 10% slow on Tuesday is an OTIF miss on Friday, but only if nobody notices until Thursday. Plants that track schedule attainment daily, by line, catch the drift while there is still time to add a shift or resequence, and the same records give sales an honest answer about what can be promised next week. Co-packers feel this hardest of all: their whole commercial relationship rides on hitting windows for volumes they do not control, which makes disciplined scheduling and live tracking a survival skill rather than a nice-to-have.
How do you build a weekly CPG schedule?
A repeatable weekly cycle looks like this:
- Pull the demand signal. Open orders, forecast, and any promotion volumes, netted against on-hand inventory that still has sellable shelf life.
- Size the runs. Convert net demand into run lengths per SKU, respecting minimum runs, batch multiples, and code-date limits.
- Assign SKUs to lines. Load each line against its real demonstrated rate, not nameplate speed, and check total hours against available hours including planned sanitation.
- Sequence within each line. Apply the transition matrix: light to dark, allergen-free to allergen, cleaning breaks where the matrix demands them.
- Check materials and people. Confirm packaging film, ingredients, and staffed positions for every run before publishing, not after the line stops.
- Lock the near window, publish, and track. Freeze the next day or two, publish to the floor, then measure schedule attainment and changeover hours to improve next week's pass.
If you are formalizing this for the first time, our free production schedule builder gives you a structured starting point for steps two through four.
What do the data and regulators say?
Reference points every CPG scheduler should know:
- The FDA recognizes nine major food allergens; sesame became the ninth on January 1, 2023 under the FASTER Act. Allergen sequencing and validated cleaning between runs are core preventive controls, which makes the run order a regulated decision, not just an efficiency one.
- Undeclared allergens are consistently among the leading causes of U.S. food recalls in FDA recall reporting, and label-product mismatches at changeover are a recurring root cause, another reason changeovers deserve schedule slack rather than heroics.
- Food manufacturing is a large employer, roughly 1.7 million people in the U.S. per Bureau of Labor Statistics industry data, and runs on thin margins, which is why schedule-driven waste (aged stock, excess changeovers, shorted orders) lands directly on plant P&L.
Where Harmony AI fits
CPG scheduling fails quietly, in the gap between the plan and what the lines actually did: a changeover that ran long, a sanitation that failed verification, a line running 80% of the rate the schedule assumed. Harmony AI is an AI-native MES, a real-time operational layer that connects lines, existing software, and the paper checklists that still govern most food plants into one live record, with no rip-and-replace. Schedulers see actual rates, actual changeover durations, and hold status as they happen, so next week's plan is built on demonstrated numbers instead of hopeful ones, and AI agents flag the drift, a run tracking short, a changeover overrunning its slot, while there is still time to resequence. Deployment is in person and white glove: Harmony AI's engineers walk your lines, map the transition rules and paperwork with your team, and stand the system up alongside what already works. See how a plant moved its paper records into one live system in the CLS case study, and compare platforms in our guide to CPG software. For sequencing mechanics in any industry, continue with our production sequencing guide.