The short answer
- A shop running many small jobs spends more time changing over than printing
- Most changeover time is preparation that could have happened earlier
- Sequencing jobs by ink and colour removes whole clean-downs
- Registration systems turn alignment from minutes into seconds
- Time it before trying to improve it
Ask a busy shop what limits its output and the answer is usually capacity. Time the day properly and the answer is usually changeover — the press is available, it is just not printing, because somebody is looking for a screen or mixing an ink that could have been ready.
Where the time goes
| Step | Could it happen earlier? |
|---|---|
| Cleaning off the previous job | Partly |
| Finding the screens | Entirely |
| Finding or mixing ink | Entirely |
| Mounting and registering | Partly |
| Setting off-contact and pressure | Partly |
| Test prints and adjustment | No |
| Locating the blanks | Entirely |
Four of those seven can be done while the press is still running the previous job, and in most shops none of them are. That is the whole opportunity: not working faster, but moving work off the press.
Changeover time is not the time it takes to set up. It is the time the press spends waiting for someone to be ready.
Preparation
Staged before the press is free
- Screens pulled, checked and racked in print order
- Ink mixed, matched to the approved reference and labelled
- Blanks counted, sorted by size and at the press
- The job sheet, proof and approval to hand
- Pallets and adhesive appropriate to the garment
- Squeegees of the right durometer, sharp
A staging area next to the press is the cheapest fix available. Everything for the next job lives there, and the changeover becomes fetching a trolley rather than a search — which also removes the commonest cause of a wrong print, which is grabbing the nearest screen.
Ink is worth preparing furthest ahead. Mixing to a brand colour while a press stands idle is expensive time, and the mix will be less careful for being rushed.
Sequencing
The order jobs run in changes how much cleaning happens between them, and cleaning is the part of changeover nobody can shorten once it is required.
Rules of thumb
- Group jobs sharing an ink colour and run them together
- Run light colours before dark to reduce contamination
- Keep water-based and plastisol runs apart where possible
- Batch jobs using the same pallet or platen size
- Run same-garment jobs consecutively
The gain is not marginal. Three jobs sharing a black ink run consecutively need one ink change instead of three, which on a busy day is a recovered hour — and a schedule built with that in mind costs nothing extra to produce.
Sequencing does have to respect due dates. The point is to sequence within what the deadlines allow, not to reorder the week around ink colours and deliver late.
Registration
On a multi-colour job, registration is usually the longest single part of setup and the part most improved by equipment.
| Approach | Effect on setup |
|---|---|
| By eye on press | Slowest; varies by operator |
| Registration marks on film | Better, still manual |
| Pre-registration jig off press | Screens arrive aligned |
| Laser or camera systems | Fastest; capital cost |
A pre-registration system is the highest-return purchase in most screen shops, because it moves registration off the press entirely and makes it repeatable between operators. It also depends on consistent screen tension — a slack screen will not hold an alignment set off press.
Standardising
Variation costs time because every variation is a decision, and decisions happen at the press.
Worth standardising
- A small set of mesh counts, chosen deliberately
- Frame sizes, so anything mounts anywhere
- Pallet sizes for common garments
- A short standard ink range, mixed only when required
- Off-contact and pressure defaults per garment type
- A written setup sequence, the same for everyone
The written sequence matters most where more than one person prints. Two operators with two habits produce two changeover times and two quality baselines, and the difference is invisible until somebody times it.
Measuring it
Record changeover as its own number, separate from run time. Without that split, a slow day looks like a busy day and there is nothing to improve against.
Where jobs are tracked in production management software, capture setup and run separately on the job record. After a few weeks the pattern is unarguable: which job types eat setup, which operators are fastest, and whether the changes you made actually did anything.
Then improve one thing at a time. Staging the next job, sequencing by ink, and pre-registration are three separate changes, and doing them together means never knowing which one earned the hour back. The rest of the production archive assumes changeover is measured rather than estimated.
Who does the changeover
In a small shop the printer does everything, which is exactly why changeover is slow: the person who should be printing is fetching screens. The first real gain usually comes from splitting the roles rather than from buying anything.
A second person staging the next job while the first runs the current one can halve effective changeover, and the second person does not need to be a printer. Pulling screens, mixing ink to a written reference, counting blanks and assembling the job folder are all tasks that can be learned in a day.
Where headcount does not allow that, the same effect comes from sequencing the day differently — staging two or three jobs during a single quiet period rather than staging each one at the moment it is needed. It is the same total work moved to a time when the press is not waiting for it.
Cross-training helps beyond the changeover itself. A shop where two people can stage a job is a shop where a sick day costs a few slow hours rather than a lost day, and the flexibility shows up in the schedule long before it shows up in the changeover figures.
Write the staging list down and keep it identical for every job. A changeover that depends on one person's memory is a changeover that gets slower whenever that person is away, and consistency across operators is worth more than the fastest individual time — the same argument as documenting the work before you hire.
