The desire for convenience, a shift in consumer tastes and preferences as well as in price/quality considerations, have all bolstered the demand for online grocery shopping. Studies have shown that optimised package images can have a huge impact on product stand out and conversion. But how do you start testing online package images?

Below you find 5 tips from our Insights Director, and more info about how many design variations to test (and types), in which context, which framework to use to evaluate and methods, and how to set up the research.

1. You only have 1-1.5s of the consumer’s attention to communicate key information on the PLP

Online grocery buyers spend typically between 15s and 20s per product list page (PLP), seeing on average 8-12 products and browsing the products only up to 1.5s. Having this in mind, you need to help customers differentiate between the key elements of the pack shot in order to minimise mistakes and increase sales. Our study has shown that Hero Images have a 5% higher visibility than Standard Package Designs on desktop (with an even greater potential on mobile, because of the smaller screen size) and can increase sales by about 15%. Further, Hero Image designs perform better in comparison to standard pack shots across most categories regarding visibility, shoppers’ interest, product type identification and purchase intent in direct preference.


2. Choose key information and attributes and test at least 2 design directions

EyeSee studies have shown that when searching for a particular SKU, up to 70% of shoppers pick the wrong one because they are overwhelmed with elements and options. While some product attributes are quite easy to grasp on the page (e.g. brand), some are more difficult. Anyone who has ever tried to differentiate a 250ml from a 400ml bottle of shampoo, solely based on their identical online pack shots, would know it is nearly impossible. Due to the size of the package shot on the PLP, an image can only communicate 1-2 messages next to the obvious brand cues. It is key to understand what really matters.

There are multiple design routes for optimising online package shots. Unilever and the University of Cambridge developed a variant that helps to recognise four key elements (brand, variant, format, and size). On the other hand, you have variants that put more emphasis on the emotional elements of the product (e.g. Spotlight hero image).


3. Testing in real-life context matters

Online platforms are dynamic environments, so testing needs to be done in a dynamic context – or at least as close as possible to it (scrollable, appropriate screen size, clicks…).

If possible, test at least two websites considering that there are different rules for their structure that affect the layout (due to a different number of columns, picture sizes can vary greatly between retailers). In case you are wondering whether to test mobile, desktop or both, primarily go with the channel through which your category is purchased the most. In case that is desktop, do include a mobile device test in a segment of the study to make your site future proof.

desktop-laptop-tablet-mobile4. Use visibility, findability and purchase interest to judge the effectiveness of the pack-shot

Studies show that up to 95% of purchase decisions are made subconsciously. Therefore, asking your shoppers to verbalize during a survey what they are not even aware of, will not get you far in understanding what triggers their purchase. A combination of conscious and subconscious measurements is strongly recommended when conducting studies like this.

Key factors and KPIs for evaluating online package performance should be:

  • Stand out: Is the package seen? Does it hold attention long enough?
  • Brand/product consistency: Is the product recognised?
  • Package engagement: Does it increase interest?
  • Sales uplift: Will it increase sales?

These variables are best measured with multiple methods. Eye tracking will provide you with insights about package stand out, response time measurement about the information processing and recognition, surveys about package engagement, and online virtual shopping with sales uplift. Here’s an overview of the KPIs and the most revealing methods:



The research project is typically a validation study, so a quantitative set up is recommended.

 5. Research setup: Use a monadic approach for each design direction (but test multiple SKUs)

So how do you approach testing? Category shoppers are recruited remotely via a panel company. The respondents are exposed to a Product list page where 5 SKUs are replaced with 1 of the new design routes. Having multiple images of the same design route in 1 PLP is also favourable. This will be most representative of the future environment, after the implementation, when multiple SKUs with the new design route would find themselves on the same PLP. Here’s a handy illustration of the research setup we recommend:



The truth is, the e-commerce grocery service is still new to online pack shots –i.e. their positioning, benefits, and challenges, respectively. Most of what we know about pain points and pack shots, through decades of research in physical stores, does not necessarily apply in the digital world. Therefore, it is crucial to assess and use behavioural insights to tap into real insight – so always test new designs and assume nothing.

Having studied thousands of e-commerce pack shots, our advice to all those looking to improve their online product images would be – less is more. Remember, the average shopper spends only 1.5s on your product! Test the packages in context for the most authentic results.

Make it as visual as possible, prioritise and set goals, and do not try to cram every bit of product info in that tiny default space in the PLP. The rules of the digital jungle are cutthroat -either you make it easy for shoppers to pull the right purchase trigger, or they will easily scroll down past your product.

Interested in testing your online pack shots? Reach out to us via