Measuring performance in packaging manufacturers: what to track and what to ignore

At the monthly review the plant head praised the night crew on the blow moulding line. Their cycle time was the fastest in the plant. Two weeks later a bottling customer sent back a truckload of jars with thin shoulders and poor neck fit. The weight checks had been skipped on nights to keep the machine running. The measure on the board had rewarded exactly what caused the return.

The overview of packaging HR says waste, speed and on-time delivery make fair goals for crews and supervisors. The detail lies in how each one is counted. GullyHR's performance management process sets the wider framework; this guide covers what packaging floors should and should not measure.

Speed is easy to read off a machine, so it drifts onto the board. But a gravure press at top speed with drifting register, a corrugator run fast with warped board, or an injection machine cycling quickly with short shots simply makes scrap sooner. Speed matters only when output passes the check. Count good output and the time lost around it, not raw speed.

• Start-up scrap per docket. Sheets, metres or kilos lost before the first approved output, measured against the job-standard for that docket.
• Make-ready minutes. From the last good output of one job to the first approved output of the next.
• Waiting for approval. Minutes a first-off sits before quality passes it. Long waits point to approver cover, not the crew.
• Output passing final check. Cartons, reels, caps or sacks that clear the final table, not those merely produced.
• Returns traced to a lot. Customer complaints linked back to machine, shift and crew through the reel label or lot slip.
• Trucks loaded on time and right. Dispatch slots met with the correct count and mix.
• Near misses reported. A rising count often means people trust the system, not that the floor got worse.

Each section has its own failure points, so the measures should name them in the floor's own words.

• Corrugator: splice failures, warp complaints, wet-end and dry-end trim waste, minutes lost at board grade changes.
• Gravure and flexo presses: cylinder or plate change minutes, register stops, shade complaints per docket.
• Blown-film extrusion: kilos sent to regrind after a gauge or layflat change, gel and die-line rejects, reels downgraded at the slitter.
• Injection and blow moulding: short shots, flash, weight and neck-fit rejects, mould change minutes.
• Die-cutting and folder-gluing: nicks, crease cracks, glue skips and cartons passing the final table.
• Woven sacks and drums: loom breakage stops, stitching rejects and drop-test failures.

• Raw speed on its own. Metres per minute or cycles per hour without a quality figure beside it.
• Individual shares of crew waste. On a machine run by three people, splitting scrap by person invites blame and hides the cause.
• Hours stayed late. Rewarding long stays rewards fatigue, and fatigue drives mistakes at the knife and the platen.
• Attitude scores. A supervisor's rating of 'attitude' says more about the supervisor than the operator.
• Daily ranking boards. Ranking crews every day pushes them to hide stops and skip checks.

1. 1

   Choose four or five measures per section

   Different sections need different measures. Corrugators, presses, moulding lines and finishing tables each get a short list from the tracked set above.

2. 2

   Fix the job-standard first

   Agree what start-up scrap and make-ready time a repeat docket usually takes, so each crew is judged against its job-standard, not against a guess.

3. 3

   Make the data easy to record

   One line per job on the shift sheet: docket, start, first approved output, scrap and stops. Avoid separate forms.

4. 4

   Link returns to the run

   Make sure every reel label, lot slip or pallet ticket carries machine, date and shift, so complaints can be traced.

5. 5

   Review monthly at the machine

   The in-charge and the crew go through the crew-scorecard together and agree one improvement for the next month.

6. 6

   Keep supervisors on the same measures

   Supervisor goals should be the sum of their crews' scorecards plus safety and training, not a separate set pulling the other way.

Machine, crew lead, shift and month.
Start-up scrap against the job-standard: better, same or worse, with the three worst dockets named.
Average make-ready minutes on repeat jobs.
First-off approval wait: longest three cases and why.
Output passing final check, against output produced.
Customer returns traced to this crew, with lot numbers.
Near misses reported, and actions taken.
One improvement agreed for next month, and who leads it.

Individuals still need reviews, but on things they control. Look at skill modules signed off, safety habits seen by the in-charge, help given to learners, attendance pattern and contribution to the crew's improvement. The rigid packaging performance page shows how an hourly operator check can feed recognition. The carton plant performance page looks at paying finishers on cartons that pass final check rather than cartons run.

Measures drift when crews learn to game them. Check a sample of shift sheets against the baler weights, the quality log and dispatch records each month. Retire any measure that nobody uses in decisions. Add one only when you know who will act on it. For scorecards and reviews shaped to your own sections, describe your floor to us.