In six months, Kobashi Foods brought scrap down by roughly 20–30% and pushed First Pass Yield (FPY%) into the low 90s. This wasn’t a lucky break—it was a measured shift to Digital Printing, better process control, and a smarter variable data workflow for their shipping and retail labels. Early wins came from practical choices, not flashy promises.
Templates and mapping tools from onlinelabels helped the team move away from manual steps. That mattered because their label runs were short, SKU counts were climbing, and shipping formats varied by marketplace. The answer wasn’t “buy everything new”; it was “make what we have work together and prove it with data.”
I led the commercial side of the project in Asia, so I’ll be candid: we argued, we threw out a few neat ideas that didn’t survive the pilot, and we stayed close to the numbers. Here’s how the story unfolded—and why the results stuck.
Company Overview and History
Kobashi Foods is a mid-sized Food & Beverage brand headquartered in Southeast Asia. Historically, they relied on Offset Printing for folding cartons and Flexographic Printing for pressure-sensitive labels. As e-commerce accelerated, SKU complexity grew from around 120 to 240 SKUs in under two years. Shipping labels had to carry GS1-compliant data and marketplace-specific fields, often changing week-to-week.
The production environment mixed Labelstock on paper and film, finished with Varnishing and Die-Cutting. Runs were mostly Short-Run or Seasonal, with batches ranging from a few hundred to a few thousand labels. The challenge wasn’t pure speed; it was staying consistent with color, registration, and variable data while avoiding long Changeover Times.
Operations kept asking a practical question—where to print shipping labels when volumes spike on a Friday afternoon without tying up the flexo press? Meanwhile, marketing had been trying to create labels in word for small promos, which worked for a single SKU but didn’t scale when dozens of SKUs needed variable fields updated from the same spreadsheet.
Quality and Consistency Issues
Color drift across substrates was the first pain point. On film-based Labelstock, ΔE would hover around 3–5; on paper-based stocks it swung even wider. Shelf inspection flagged inconsistencies in brand reds and greens, which hurt perceived quality. Reject rates sat around 7–9%, partly due to registration and partly due to variable data mismatches on shipping labels.
Humidity didn’t help. During the rainy season, adhesive behavior changed and lamination tension fluctuated. Some rolls curled post-varnish, and cartons with matched labels looked off under store lighting. This wasn’t a catastrophic failure; it was the kind of day-to-day friction that drains capacity and confidence.
Data workflows were messy too. The team literally asked, “how to print labels from excel spreadsheet without reformatting every time?” Mis-mapped fields led to manual relabeling and wasted rolls. A few times, shipping barcodes scanned fine but product data for marketplaces didn’t match, leading to returns that were preventable with tighter templates.
Solution Design and Configuration
We selected a Digital Printing path with UV-LED Printing for labels. The combination offered stable curing on mixed substrates and shortened setup compared to Flexographic Printing for frequent changeovers. For embellishment, we kept it pragmatic: Varnishing for scuff resistance and clean Die-Cutting. Variable Data was a core requirement—GS1 barcodes and DataMatrix had to render within ISO/IEC 18004 tolerances.
On the workflow side, we standardized a file prep process and introduced mapping using an onlinelabels template approach. Instead of relying on teams to create labels in word for one-off jobs, we established a VDP pipeline: spreadsheet fields are validated, then merged into print-ready PDFs. This removed the weekly debate about fonts and layout drift and made changeovers more predictable.
For export SKUs, Canada-bound shipments needed bilingual elements and specific barcode placement. The team referenced onlinelabels canada resources to validate content blocks and spacing on Labelstock. That small step helped align with retailer expectations without spinning cycles on design tweaks late in the week.
Pilot Production and Validation
We ran a four-week pilot on mixed Labelstock: paper for local retail, PET film for chilled items. Under a basic ISO 12647 color control, ΔE moved toward ~1–2 on brand-critical colors. FPY% rose into the ~90–94% range on three of four pilot weeks. Changeovers dropped from roughly 40–60 minutes to ~15–25 minutes with standardized print-queue rules and preflight checks.
We hit a snag the first week: slight curling on a laminated batch for chilled products. We traced it to tension settings and lamination thickness; switching to a thinner laminate and adjusting unwind tension fixed it. Another lesson: water-based Ink on specific films needed longer dwell time; UV Ink was steadier in our conditions, though we remain mindful of Low-Migration Ink requirements for food contact zones and label placement.
Quantitative Results and Metrics
Six-month results stayed consistent enough to build confidence. Waste rate came down by roughly 20–30%. Throughput per shift grew around 12–18% depending on SKU mix. FPY% stabilized mostly in the low 90s, with occasional dips when SKUs switched to unfamiliar substrates. Energy per pack (kWh/pack) tracked about 8–12% lower versus the prior setup, a bonus discovered during utility reviews.
Color accuracy and inspection data held up: ΔE overall clustered around ~1–2 on critical hues. Registration issues were still present in small pockets, especially on high-shrink film orders, but the defect rate narrowed. Changeover Time stayed closer to ~15–25 minutes when the team followed the VDP checklist and preflight routine.
The financial side wasn’t a moonshot. Based on volume and material mix, the Payback Period landed near 14–18 months. We didn’t count fringe savings; we only modeled scrap, press time, labor, and rework returns. One caveat: marketplace promos spike unpredictably, so maintaining FPY% above 90% requires discipline in template control and data validation—exactly where onlinelabels template mapping kept the team honest. And yes, we’ve built in a playbook for future bilingual lots using references from onlinelabels canada.
