When Sticker Giant first heard from Fjordgood, the email subject line was stark: 'Our labels are lying to us.' Lena Brandt, operations manager at the Berlin frozen food startup, had watched barcodes curl off meal pouches in a -18°C warehouse. The food inside was fine. The labels were not.
I called her back within the hour. Lena didn't want a sales pitch; she wanted to know what kind of freezer labels could survive six months at -18°C without peeling, smearing, or losing scannability. It was the right question, and it deserved a technical answer.
This is the story of that answer. It starts with a Google Sheets project gone wrong and ends with custom shipping labels that actually scan in the rain. It also proves that the cheapest label is rarely the least expensive one.
From a Shared Kitchen to a Cold Chain
Fjordgood began in 2019, when two friends started cooking healthy dinners in a shared Berlin kitchen. What started as a favor for friends snowballed into a real business: by 2024, they were producing 12,000 meal pouches a week in a contract facility outside the city. Lena handled procurement, labeling, and quality—basically every job that didn't require a chef's hat.
The labels were an afterthought at first. They bought a laser printer and searched for how to make labels in google sheets, using a template with the brand's logo, allergen text, and best-before dates. It worked fine for jars of granola. It did not work in a warehouse where every surface was cold enough to bite through a jacket.
Within a month, the first complaints arrived. FedEx found a torn label and refused a delivery. A retail partner's scanner couldn't read a barcode. Lena started weighing every returned pouch and putting the peeled labels in a folder on her phone. The folder had 41 photos by the time she wrote to us.
The Hidden Cost of a Label That Peels
'A label failure doesn't stay on the label,' Lena said during our first video call. On a frozen meal pouch, the label carries heating instructions, allergen warnings, ingredient declarations, and lot traceability. If it curls off the pack, the product can't be sold—and if it carries the wrong batch code, it can trigger a recall. Fjordgood lost 1,400 pouches after one retailer couldn't scan a sleeve of best-before dates. Nobody shouted. The silence was worse.
Lena still laughs about a half-finished search she typed in those weeks: 'giant college sticker price isnt most.' She didn't mean anything about college or stickers. She meant that the number printed on a label supplier's price list wasn't the number that mattered. The cost of waste, rework, and a damaged retail relationship was several orders of magnitude bigger.
That insight became the turning point. Once she framed the problem as total landed cost—not cost per label—Fjordgood was ready to test a different approach.
Why the Answer Wasn't a Printer Template
We sent Fjordgood a kit of candidate label stocks. Each sample was printed with digital ink, then frozen and rubbed with a damp cloth. The winner combined a 60-micron polypropylene face stock, a water-based adhesive rated for -20°C, and a matte topcoat that didn't smudge under condensation. It wasn't the cheapest sample. It was the one that worked.
The broader solution used two different products. For the meal pouches, we ran short-run digital freezer labels with variable date codes. For the outer cartons, we made custom shipping labels printed on the same polypropylene film, with a stronger adhesive to survive courier handling and rain. Minimum order sizes were low enough that Lena didn't need to warehouse thousands of obsolete labels.
There was one awkward moment: the matte finish we tested looked so good in the photos that Lena almost chose it based on aesthetics alone. Then the freezer test showed it peeled after 24 hours in fluctuating temperatures. Aesthetics matter, but not more than adhesion.
The Rollout: Fast, But Not Perfect
The rollout began with a sample run of 500 pouches. We set the digital press to transfer batch and best-before data onto each label in real time. On the packing line, labels moved through an automatic applicator at 45 pouches per minute without any jams. Lena's team did not have to re-print a single batch.
Then the spiral freezer introduced an unexpected variable. Cold air created static cling, and some labels shifted before the adhesive fully set. A humidity controller near the applicator fixed it, along with a static-dissipative version of the labelstock. The issue wasn't a failure of the label itself, but a reminder that real production is never exactly like a test.
Six Months Later: What Stuck
Six months after the switch, Fjordgood's label failure rate fell from around 6% to under 0.5%. The team stopped hand-re-labeling, which saved roughly 12 hours a week. And the grocery chain that had quarantined their order returned with a larger contract after two audit-free deliveries.
'I don't miss the DIY days,' Lena says. 'Now we order freezer labels in short runs whenever we need them, and Sticker Giant delivers in about five days.' She paused. 'And I no longer stay up at night thinking about how to make labels in google sheets.'
For me, the case underscores why labels deserve a seat at the design table. Adhesive, face stock, ink, and barcode all work as one system. When it clicks, you notice nothing. When it fails, the whole product story falls apart. Fjordgood's label system now does exactly what a label should: stay quiet, stay put, and make sure the food gets eaten.
