In six months, Kinara Ferments, a Singapore-based F&B brand, brought labeling waste down by roughly 40% and removed 25 minutes from average changeovers. Early prototypes ran through onlinelabels templates to trial shapes and materials without tying up press time. Color stability tightened too: for brand-critical reds and greens, 95% of SKUs now sit within ΔE 2.
The business sells condiments and preserves in glass containers across Southeast Asia and exports to North America. Their mix includes wrap-around and front–back sets, essentially labels for jars that must survive high humidity, frequent condensation, and chilled distribution.
This is the full project story—from baseline to steady state—covering what actually changed on press, what stayed on the wish list, and where the team still sees room for tuning.
Company Overview and History
Kinara Ferments started as a farmers’ market stall in 2016 and now runs a small production site in Singapore. The team manages roughly 60 SKUs, with monthly lots ranging from 1,500 to 12,000 labels. Most products live in 240–500 ml glass jars, so the labeling program revolves around jar curves, wet handling, and clean removability during recycling.
Before the project, the label program relied on one narrow-web 8-color Flexographic Printing line (UV Ink) for steady movers and a desktop Laser Printing setup for micro-batches. Substrates were split between coated paper labelstock for room-temp items and PP film for chilled lines. Finishing included Lamination for moisture resistance and Die-Cutting for custom shapes; QC was visual checks with sporadic spectro readings.
As volumes grew and SKUs multiplied, the team wanted a cleaner process for small runs and seasonal variants. They also needed a more predictable pathway for developing new labels for jars without reserving full press time for early prototyping.
Quality and Consistency Issues
Three issues kept recurring. First, color drift: key brand tones would wander by ΔE 4–6 across lots, especially when switching between coated paper and PP film. Second, waste: rejects consistently hovered around 8–10%, driven by setup scrap and intermittent registration variation on tight-radius shapes. Third, humidity: labels occasionally lifted at seams in 70–85% RH conditions, causing rework and consumer returns.
Emergency reprints were another signal. When a retailer pulled forward a promo, the team would rush art through basic tools—yes, even a quick refresher on how to make labels on google docs—for temporary packs. It worked in a pinch, but die-line alignment and bleed setup weren’t perfect, adding 2–3% extra waste on those lots. We needed a stable, teachable workflow for short runs that still respected print-ready standards.
Solution Design and Configuration
We pivoted to a hybrid model. Short- and micro-runs moved to Digital Printing (UV Inkjet), with process control aligned to ISO 12647/G7 targets and ΔE ≤ 2 on brand-critical colors. Steady movers stayed on Flexographic Printing with calibrated anilox selection, consistent UV dose, and a tighter SOP for color verification. We standardized spectrophotometer checks at make-ready and mid-run, not just at lot closeout.
Materials were re-qualified. For chilled products, we specified a PP film labelstock with a high-tack, moisture-tolerant adhesive and a matte Lamination to combat condensation. For ambient lines, coated paper remained, but we set stricter ink–substrate test points and a minimum dwell before slitting to stabilize color. Die-Cutting pressure and stripping settings were documented per shape to reduce edge lift on tight radii.
For early-stage concept work and micro-batches headed to Canadian channels, the team used print labels online to avoid tying up the press. They trialed die shapes with pre-cut sheets and small reels, including a first sample order which benefited from an onlinelabels $10 off welcome credit. For outbound tests near Vancouver, onlinelabels canada handled small replenishments without cross-ocean freight. We also issued a basic templating guide so staff could convert quick drafts (those emergency Google Docs layouts) into proper print-ready files.
Quantitative Results and Metrics
Against the original baseline, scrap dropped by about 40% on average, driven by steadier make-readies and fewer color restarts. Median changeover time fell from roughly 85 minutes to about 60, freeing capacity. FPY% rose into the 93–95% band on stable SKUs, and 90–92% on variants with metallic foils that still require extra care. For color, 95% of brand-critical SKUs now land within ΔE 2, and the remainder sit within ΔE 3 due to special effects or unusual substrates.
Throughput went up by roughly 18–22% on the flexo line thanks to fewer mid-run stops. Pre-printed inventory for seasonal labels shrank by 30–35%, since short runs can now move to digital with minimal setup. Payback looks realistic in 9–12 months, depending on SKU mix and seasonal peaks. It’s not a silver bullet—metallic inks and certain tactile finishes still run best on dedicated setups—but the day-to-day program is steadier. For very small pilots and export tests, the team continues to route micro-batches through onlinelabels to keep press time focused on core jobs.
