Experimentation / CRO

Experimentation and Testing Programs acknowledge that the future is uncertain. These programs focus on getting better data to product and marketing teams to make better decisions.

Research & Strategy

We believe that research is an integral part of experimentation. Our research projects aim to identify optimization opportunities by uncovering what really matters to your website users and customers.

Data and Analytics

90% of the analytics setups we’ve seen are critically flawed. Our data analytics audit services give you the confidence to make better decisions with data you can trust.

How Product Image Size Affects Attention and Engagement [Original Research]

Bigger images should make you pay more attention to them, right? Well, maybe. We decided to test that assumption in regards to eCommerce product pages. In this study, we explore how elements of a product page affect users' visual perception and perceived product value.

This experiment looks at how viewers perceive a product page when the product image size changes.

How Product Image Size Affects Attention and Engagement [Original Research]

Results summary

  • The spec-driven product (hard drive) shows a pattern of increased visual attention with increased image size.
  • The experience–or design-driven product (men’s dress shirt) shows a pattern of decreased visual attention with increased image size.

How do I apply this research?

  • If you have a spec-driven product (e.g.hard drive, camera, printer, software, etc), test increasing image size to capture more visual attention.
  • Combined with results from part 1 of our eCommerce product page study, this could mean that people value the product more when the image is larger, resulting in increased visual attention.

Full Study Details - Visual Perception & Image Size

This study tested the hypothesis that a larger product image attracts viewers' attention to the image more compared with a smaller image. (After normalizing for image size.) This study was conducted using eye-tracking.

For each combination of product type (n = 3; dress shirt, headphones, hard drive) and size variation (n = 2; large and small), we had at least 44 people view the page with the task of assessing the product with intention to purchase. A single participant would view the 3 products within a single image size variation. So, all large products or all small products.

Participants for each format treatment
Participants for each size variation.

Treatment Heatmap Results

Aggregated heatmap gifs for all treatments.

Statistics and Results Summary

Let's see what the data tells us. To assess the relative importance of the images, we measured the average time fixating on the image. Here's a table of the raw data for the treatments:

Summary data for time fixating on the image
Summary data for time spent fixating on the image

And, for a better view of the data, below is a histogram of the average time fixating. In t-tests between the size variations for each product, we note that there are no significant differences at the conservative alpha of 0.05. However, there are some trends that may help refine our hypotheses and provide support for why our study had the results that it did.

Histogram of the average time fixating on the product for each treatment. Arrows indicate direction of effect of the smaller image
Histogram of the average time fixating on the product for each treatment. Arrows indicate direction of effect of the smaller image.

The interesting thing to note here is that we see a higher average time fixating for only the small version of the shirt, so not the headphones or hard drive. People spend more time looking at the small version of the shirt, but the larger version of both the headphones and hard drive.

We also ran Analyses of Variance (ANOVA) tests among the product types for the small and large variations. This was to simply see if there were interactions between the amount of time people fixated among the products.The differences are easily visualized in the histogram above, so we'll simply report here that, for the large treatments, we did see significant differences between:

  • The shirt and headphones.
  • The hard drive and headphones.

But we did not see a significant difference between the shirt and hard drive.What does this mean?

It means that the headphones were more visually engaging than the hard drive or the shirt, so it kept people's attention longer on average... but only when using the large image.

The small variations resulted in no significant differences among the three product classes. With the above results, combined with the value perception data we have from Part 1 of this study, we have an interesting story here. A common thought in e-commerce product page setup is that large images are better, more visually engaging and offer an improved experience for the consumer. What we're finding here, however, is that this might not always hold true across all types of product classes.


This study tested only one product in each of the extremes of the 'experience-search' product classification regime. Stronger support would come from testing across multiple (the more, the better) products in each of the classes. For example, the result we saw for the shirt, both in this study and in Part 1, could have been influenced by consumers' online shopping biases towards a wearable.

That means they don't necessarily draw conclusions from the image itself (except for color, pattern, etc.) for fit and feel. They might actually draw conclusions from the gestalt quality of the site. Thus, when more white space is present (with the smaller image), a higher value is perceived (per Part 1 results).

Our sample size may not have been sufficient to detect an effect. Additional participants may have improved the statistics and potentially led to significance at the alpha=0.05 level for the difference between sizes for both shirt and hard drive products. A notable limitation in this work is that visual patterns don't always correlate to purchasing patterns. We therefore do not try and place 'value' judgements on the conclusions with regards to buyers' purchasing behavior. Rather, the conclusions here are intended to aid optimizers in terms of what to test.


We initially thought that larger images would increase the attention that users paid to it. But that wasn't exactly true. Rather, with experience-driven product (men's shirt), smaller images produce more attention (visual engagement). But, for our spec-driven product (hard drive), larger images resulted in more visual attention.

Related Posts

Who's currently reading The Experimental Revolution?