← Back to Our Work
Print Operations · Workflow Design
A major Canadian financial institution was producing 5,000 pieces per day on a Pitney Bowes Epic inserter with completely random variable page counts — forcing the inserter to constantly change speed and settings with every piece, and requiring hundreds of manual pull extractions per run. The operation was running far below its capable throughput on every single run. The solution required no new equipment and no capital investment — only a redesign of the pre-production data workflow.
The job file arrived with completely random variable page counts — 1, 5, 4, 3, 2, 1, 1, 5, 4, 3, 2, 2, 1... — forcing the Pitney Bowes Epic inserter to constantly adjust speed and settings for every single piece. The machine never reached or held optimal throughput because the page count changed with every job that ran through it.
Compounding the throughput problem, the client required hundreds of specific piece extractions — pulls — per run. Operators were stopping the production line mid-job to manually locate and pull individual pieces at the finish stage — the most disruptive and costly point in the workflow. The cumulative effect was significant daily throughput loss on every run.
Mike Preczner redesigned the pre-production data workflow to sort and stream the job file before it ever reached the inserter floor — solving both the throughput problem and the pull problem in a single upfront process.
Page-Count Stream Sorting: The full 5,000-piece job file was pre-sorted upfront into five clean, sequential streams by page count. The inserter runs each stream independently at its optimal speed setting for that page count — constant speed, consistent makeready, maximum throughput maintained throughout each stream. No more speed changes between pieces.
Automated Pull File Extraction: All client-requested pulls were extracted from the data file upfront during the streaming process — before the job reached the floor. Pulls were compiled into a separate, dedicated file handled digitally at the front end. Hundreds of individual piece extractions that had previously required manual line stops were resolved in a single automated pre-production step. The inserter ran the full job uninterrupted.
| Metric | Before | After |
|---|---|---|
| Job file sequence | Random — 1,5,4,3,2,1... | Pre-sorted into 5 clean streams |
| Inserter speed | Constantly changing — every piece | Consistent — optimal per stream |
| Inserter throughput | Far below capable rate | Maximum throughput per stream |
| Pull process | Hundreds of manual mid-job stops | Automated upfront extraction |
| Pull timing | Finish stage — most disruptive | Pre-production — zero disruption |
| Line stops for pulls | Hundreds per run | Zero |
| Capital investment required | Yes — assumed | None — data workflow solution only |
The transformation required no new equipment and no capital investment. The solution was purely a data and workflow intelligence fix — redesigning how the job file was prepared before it reached the floor. A 5,000-piece daily run that had been chronically underperforming was converted into a clean, predictable, high-throughput operation running optimally on every stream, every day. Directly transferable to any high-volume variable page-count insertion operation facing the same throughput ceiling.
If this engagement reflects a challenge your operation is facing, let’s talk. Swift Supply Solutions delivers practitioner-led consulting and fractional executive leadership for organizations ready to move. A conversation costs nothing.
Book a Conversation →