Do you know which key information your brand needs to communicate on each SKU to attract e-shoppers? Most companies understand what their products’ purchase drivers are, but the majority haven’t tested whether these are reflected in their e-Content.

That’s what we did for the Spanish leading honey brand, Granja San Francisco. We analyzed their brand position, detected each SKU’s main purchase drivers and tested whether the packaging was aligned to this messaging or if it was failing to communicate properly.


The AI tracking test

Artificial Intelligence (AI) tracking technology emulates the behaviour of the human eye to analyze e-Content the way a person browsing the website would. This shows us whether the content is communicating what it needs to communicate, and whether or not the messaging is clear. 

We used this technology throughout the three stages of the project: 

  • During the audit stage to find out what the original packaging was failing to communicate;
  • During the creative process to help the design team create the assets;
  • In the final test to check that the resulting e-content aligns with the desired messaging.


The AI tracking analysis is based on three elements:

-CLARITY: how well key elements stand out within the context of the image.

When first processing new visual information, humans will only perceive up to 5 salient regions AI tracking measures the complexity of the design on an index of 0-100. The median is 50. Tracking above 50 suggests that the average brain will struggle to decipher the overall content.

-DIGESTIBILITY: whether a clear visual hierarchy is present in the design.

On average, web page users only have time to read up to 28% of words per visit. AI tracking creates numerical sequencing on opacity maps to indicate the sequence that people look at the key content regions and the order that they are seen (or skipped entirely). Analysis is based on the metric scores for the regions.

-GOAL ALIGNMENT: whether viewer attention is aligned with the goals.

A design may have a high degree of digestibility and clarity, but it could still be misaligned with the brand’s overall goals. AI tracking uses a traffic light system to indicate how well critical sub-level information, such as labelling information, is being delivered. Typically this is used to compare the intended delivery of product messaging hierarchy.


AI tracking test of Granja San Francisco packaging

We started by defining which key information the packaging should communicate, and then we tested if the actual packaging was aligned or not. On the plus-side the design was uncluttered and the masterbrand was clear, but the product sub-brands and enhanced flavours were underperforming.


A new e-Packaging design based on data and performance

The new e-Packaging design, with improved product lighting, has enhanced the overall readability of the masterbrand, the product brand and the “0% Goteo” flash. Additionally, the “Vitamina C” flash has gained higher visibility compared with the unoptimised ePack.

We tested the new e-Packaging design again, and the results showed an improved performance in the three elements of analysis. We have improved the clarity of the key elements, the new e-Packaging is easier to digest which means the e-content hierarchy is reflected better by the packaging, and the goal alignment shows more coherent features.

This successful e-content optimization project is based on data with the objective of creating better performing basic assets like product images.