In this instalment of Datawatch, we explore how brands are using consumer insights to design the perfect packaging for their products.  

It’s no secret that big data can be used to generate insights into almost every stage of the supply chain. It’s used by supply chain managers to optimise every part of a product’s production and distribution, from the prototype to the point where it hits the shelf. And now, big data is even being used to guide how we design packaging. 

Using data science techniques, brands can study consumer behaviour to see how specific design elements affect the sales of their products – and provide their designers with valuable insights to guide their design process. Here’s how.  

The influence of product packaging

Dozens of different factors affect consumers’ purchasing decisions while they’re shopping – from product availability to the ease of the purchase – but few have as big an impact as packaging design. It’s the first impression the consumer has of a product, and it offers a sense of the idea they’re buying into. 

In fact, research shows  that 72% of US consumers agree that the design of a product’s packaging influences their purchasing decisions, and 67% said the same for the materials used to package a product.

In the past, packaging designers have relied on common trends to guide the designs of specific products. For example, certain colour patterns convey different messages – black, silver and gold are used for luxury products, while white is used for cleaning products. 

But insights like these are very high level, and they’re often not based on hard data. Now, many product designers are using the big data collected by their brands on consumer behaviours to generate more granular insights into the designs and packing types that customers are most responsive to.

Crucial insights on customer psychology

Packaging Strategies magazine offers a valuable example of how granular these design insights can be. The magazine mined its database of more than 20,000 design systems to draw links between KPIs and purchases from shelves to identify how specific elements of packaging design affect product performance.

Using its machine learning model, the team coded changes to individual design elements such as logos, colour, visuals, structure, and claims, to compare proposed designs to the product’s current packaging. And the model revealed some interesting patterns:

  • Changes to brand identity, such as logos and variant descriptors, were more strongly related to sales declines – suggesting the elements appeared to confuse shoppers and create hesitation.
  • Changes to pack colour, visuals, and shape were more strongly connected to improvements in visibility of the product on the shelf and an increase in purchases – suggesting design changes need to be significant enough to be noticeable from several feet away. 
  • New design systems that include both graphic and structural changes and new claims were more likely to drive sales gains – suggesting an important link between design and messaging when approaching packaging re-stages. 

While these patterns can’t be used to guarantee the success of a product’s sales, they’re incredibly useful for briefing marketers and packaging designers with more data-driven guidance.  

Putting design patterns into practice

While the study by Packaging Strategies magazine just tested what’s possible with consumer data, there are already brands out there putting these insights to use.

Recently, we worked with one of the world’s leading sugar refining companies – to help the brand identify how its customers respond to product claims on its packaging. Working with The Smart Cube, the company surveyed its customers to capture data on their preference of the brand’s six key product claims – including claims such as “100% Natural Cane Sugar”, “US Food Safe”, and “No Artificial Colours, Flavours, or Preservatives”.

Using this information, The Smart Cube modelled the data to identify how these claims affected four KPIs for both bakers and non-bakers, including the likelihood to buy, appeal, relevance, and believability. This modelling allowed the brand to generate specific insights on which claims it should include on its packaging, how prominent they should be, and which claims aren’t worth including. 

For example, “100% Natural Cane Sugar” and “Pure Cane Sugar” were both terms that prompted the highest likelihood to buy across bakers and non-bakers, closely followed by “No Artificial Colours, Flavours, or Preservatives” and “100% Natural”. Other findings included specific claims such as “Adds Flavour to Food” and “USDA Organic” prompted even greater likelihood to buy from its baker audience compared to non-bakers. 

Equipped with these insights, our client can now offer data-driven insights to its packaging designers when they’re designing new products – including specific guidance for products targeted towards bakers and non-bakers. 

Limitless potential for data-led packaging design

As brands continue to capture more customer data related to purchasing psychology, they’ll be able to generate even more accurate insights to guide how their packaging is designed. Whether it’s through customer surveys like the one conducted by the sugar refining company, or through innovative new packaging formats such as IoT-enabled packaging and active packaging that uses QR codes, the potential for data-led packaging design is limitless. 

Here at the Smart Cube, we offer bespoke, end-to-end analytics capabilities from data engineering through to reporting and visualisation – to help equip organisations across every industry with the intelligence they need. 

To read about some of the ways we’re helping our clients, or learn how we can help you achieve your own business goals, visit here