In 2017, Nutella’s Italian campaign produced seven million jars with distinct, colorful labels by combining patterns and color variations, then printing each design with a unique identifier. The campaign was called Nutella Unica. Reports describe an algorithmic design-and-printing process—not machine learning or generative AI.
How did Nutella make seven million different labels?
The campaign started with a set of visual patterns and color combinations. Software applied transformations such as rotating, zooming and flipping pattern areas, then combined them to produce many different label graphics. The Drum’s 2017 campaign account says the designs drew on dozens of patterns and thousands of color combinations; Wired Italia reported that the development used twelve patterns and took two years. These descriptions come from campaign and press coverage, not a technical specification of the algorithm.
From patterns to printed labels
- Pattern and color inputs: Designers prepared patterns and color variations for the label artwork.
- Automated variations: The process transformed and combined those inputs—using operations including rotation, zooming and flipping—to create distinct graphics.
- Digital label production: A label-industry report identifies HP Smartstream Mosaic as the software used to automate design variation and pass the output to digital printing equipment. This is secondary trade reporting, rather than direct documentation from the software maker or campaign team.
- Identification: Campaign coverage says the printed labels carried unique ID codes; contemporaneous Italian coverage describes the jars as serial-numbered.
Those identifiers distinguish jars in the campaign reporting; the sources do not establish that they provided authentication or anti-counterfeiting protection.
Was the Nutella Unica campaign really AI?
The available descriptions support calling it algorithmic design: software generated variations by transforming and recombining patterns and colors. They do not establish that the system used machine learning, trained a model or relied on generative AI. Calling it AI would go beyond what the campaign and trade reports explain.
Recommended Free Tools
#1 Best Overall
What was the campaign, and how large was it?
Nutella Unica was an Italian campaign launched in 2017 around the idea that individually designed packaging could make a mass-market jar feel personal. The Drum’s campaign case coverage, which credits Ogilvy Italy, reports seven million different jars and says they sold out in one month. That sell-out is a campaign-reported outcome, not an independently audited sales figure.
Wired Italia reported that the launch included three jar sizes: 630 g, 825 g and 1 kg. Those are historical launch details, not a statement about Nutella’s current product range. Italian coverage also describes television, radio, web and social media as part of the launch.
Rank #2
- Ideal for Gifting
- Ideal for a bookworm
- Compact for travelling
What the seven million labels show—and what they do not
The campaign is a practical example of how automated variation can scale packaging design: a finite set of inputs and transformations can yield a very large run of distinct graphics for digital printing. It does not, by itself, reveal the algorithm’s code, prove that every possible design was visually unique, or document a particular degree of randomness. The sources describe the campaign’s output and broad workflow, not its technical implementation or performance metrics.
The jars belong to a 2017 Italian campaign and should not be assumed to be currently available. The reported scale and sell-out are useful context, but neither is backed by an audited sales record in the cited campaign coverage.
Quick Recap
Best Value
- It can be a gift option
- Comes with secure packaging
- Helpful in various ways
Rank #4
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
