Product images are designed to sell.
They introduce the product, communicate key information and help shoppers decide whether an item is worth exploring further. But the way those images are created and reviewed is often very different from the way shoppers actually experience them.
A creative team may review an image on a large, high-resolution monitor.
A shopper may encounter the same image as a small thumbnail on a smartphone, alongside several competing products.
That difference in viewing conditions can have a significant impact on how easily critical information can be read.
A product image can look perfectly clear during the design process and still become difficult to read when displayed at the size and scale used on a digital shelf.
This is why product image testing needs to consider the environment in which shoppers actually encounter the content.
The question is not simply:
“Does the image look clear?”
It is:
“Can the critical text within the image potentially be read at a glance when displayed in a mobile shopping environment?”
Why Product Images Need Mobile Readability Testing
Mobile commerce has changed the way shoppers interact with product imagery.
Consumers increasingly browse products on smartphones, where screen space is limited and product images are often displayed at relatively small sizes.
On a product listing page, shoppers may scroll rapidly through multiple products, comparing packaging, product names, variants, sizes and benefits.
They rarely have the time or inclination to stop and study every image.
This creates a particular challenge for brands.
The information contained within a product image needs to remain sufficiently readable when the image is encountered at the scale and viewing conditions that shoppers actually experience.
A product image might contain:
- Brand
- Product type
- Variant
- Size or pack count
- Key benefit
If critical text becomes difficult to distinguish when the image is reduced to mobile size, the image may fail to communicate important information even if the overall design remains visually attractive.
Mobile readability testing provides a way to identify these potential problems before assets are published.
What Should You Test in a Product Image?
Product image testing should look beyond whether the image appears visually appealing.
Several measurable characteristics can influence the potential readability of text within an image.
1. Text Character size
Character size is one of the most important considerations.
When an image is reduced to thumbnail size, the apparent height of every character decreases.
Text that looks comfortably readable on a desktop monitor may therefore become difficult to distinguish on a smartphone.
Testing character size provides an objective way to identify text that may struggle to remain readable at its intended display size.
2. Contrast
Text also needs to be sufficiently distinguishable from its background.
Poor contrast can make even relatively large characters difficult to distinguish, particularly when text is placed over photography, packaging or complex visual elements.
RHINO uses the Advanced Perceptual Contrast Algorithm (APCA) alongside other image measurements to assess text contrast.
Contrast should not be considered independently from character size. A small character with strong contrast may present a different readability challenge from a larger character with weak contrast.
3. Font weight or Stroke Thickness
Font weight is another characteristic that can influence potential readability.
Text that is extremely light may become harder to distinguish when displayed at a smaller size, while heavier text can present a stronger visual form.
Testing font weight alongside character size and contrast provides a more complete assessment than relying on a single design rule.
Even when the foreground colour remains unchanged, variations in the background can alter how clearly the text is distinguished.
4. Text presentation
Product images can contain uppercase text, lowercase text and numerals.
These different forms of text can perform differently under glance-reading conditions.
A robust testing approach therefore considers whether text is presented as uppercase, lowercase or numerals rather than treating all characters as equivalent.
Why reviewing at Mobile Size Matters
One of the most common mistakes in product image approval is evaluating the asset at the wrong scale.
A designer might open a 2000-pixel image on a large monitor and see text that appears perfectly readable.
The shopper does not necessarily experience the image that way.
The same asset may appear as a much smaller representation within a product listing.
As the image shrinks:
Characters become smaller.
Fine details become harder to distinguish.
Contrast problems can become more noticeable.
Text can compete with other visual elements.
Critical information can become difficult to recognise quickly.
This means that reviewing an image at its original design scale does not necessarily tell you how it will perform when encountered by a shopper.
The viewing condition matters.
Testing therefore needs to consider the intended display environment rather than relying solely on the size at which the creative team happens to review the original asset.
Manual vs Automated Product Image Testing
Manual review has an important role in creative production.
Designers and brand teams need to assess whether an image is visually compelling, consistent with brand guidelines and appropriate for the product.
But manual review becomes less reliable when the question is specifically about potential readability at scale.
Manual testing
A manual process might involve:
- Opening the image.
- Reducing it to an approximate mobile size.
- Inspecting the critical text.
- Checking whether it appears sufficiently clear.
- Making subjective recommendations.
- Repeating the process for every asset.
This can work for a small number of images.
However, it becomes difficult to maintain consistency across hundreds or thousands of assets.
Different reviewers may reach different conclusions.
The same reviewer may also assess an image differently depending on familiarity with the content.
Automated testing
RHINO AI-powered image analysis provides a more consistent approach.
Instead of relying entirely on visual judgement, automated systems can identify text within an image and measure characteristics that influence potential readability.
RHINO uses Optical Character Recognition (OCR) to detect text within product imagery.
It can then analyse characteristics including:
- Character size
- Font weight
- Contrast testing of text vs. background
- Uppercase, lowercase or numerals
These measurements can be evaluated consistently across large image libraries.
The purpose is not to replace creative review.
It is to provide objective evidence that complements creative judgement.
How RHINO AI Can Analyse Product Images
Automated product image testing typically begins by identifying the text contained within an image.
This is where Optical Character Recognition (OCR) plays an important role.
OCR allows text within imagery to be detected automatically rather than requiring someone to manually identify every word.
Once text has been identified, the image can be assessed against measurable characteristics.
For example:
Step 1: Detect the text
OCR identifies words and short phrases within the image.
Step 2: Measure character size
The system assesses the size of the characters within the image.
Step 3: Analyse text characteristics
The assessment can consider font weight, foreground colour, background colour and text presentation.
Step 4: Assess contrast
The relationship between foreground and background can be evaluated using APCA.
Step 5: Interpret the measurements
The measured characteristics can then be considered against relevant experimental evidence from the Cambridge Contrast And Size Tests (CamCAST).
This moves the process from simply asking whether text “looks readable” towards an assessment grounded in measurable characteristics and experimental evidence.
The Role of APCA in Product Image Testing
Contrast is an important part of product image readability.
Text can become difficult to distinguish when it is placed over backgrounds with insufficient contrast or significant visual complexity.
RHINO uses the Advanced Perceptual Contrast Algorithm (APCA) as part of its image assessment methodology.
APCA provides a perceptual approach to evaluating the relationship between foreground and background colours.
This is particularly relevant to digital commerce imagery, where text may appear over:
- Product photography
- Packaging
- Coloured backgrounds
- Gradients
- Lifestyle imagery
- Other visually complex areas
However, contrast should not be treated as a standalone measure of glance readability.
Character size, font weight and text presentation also contribute to the potential readability of the text.
For a deeper explanation of APCA and its role in product imagery, see:
What Is APCA Contrast and Why Does It Matter for Product Images?
The Role of CamCAST and PGI
Objective product image testing becomes more meaningful when measurements can be interpreted using experimental evidence.
The Cambridge Contrast And Size Tests (CamCAST) provide experimental data relating factors such as character size, contrast, font weight and text presentation to glance readability under controlled conditions.
RHINO uses the relevant CamCAST evidence when assessing text within digital imagery.
The methodology can also express potential readability through Potential Glance Inclusion (PGI).
PGI is an estimate of the proportion of UK adults who could potentially read specified handheld text at a glance under the conditions represented by the calculation.
Importantly, PGI measures potential readability, not comprehension.
It does not guarantee that someone will notice, successfully read, understand or act on the information.
For a detailed explanation of PGI and how it relates to CamCAST and RHINO:
What Is Potential Glance Inclusion (PGI) and How Is It Calculated?
A Practical Product Image Testing Workflow
For brands managing large digital image libraries, product image testing can be incorporated into the existing content workflow.
Step 1: Identify the critical information
Determine which words within the image are important to product identification or understanding.
This could include the brand, product type, variant, size or key benefit.
Step 2: Consider the intended display environment
Assess the image in the context in which shoppers will encounter it.
For mobile commerce, this means considering how the image appears when reduced to an appropriate mobile or thumbnail display size.
Step 3: Detect the text
Use OCR or another image-analysis method to identify text within the image.
Step 4: Measure the relevant characteristics
Assess:
- Character size
- Font weight
- Foreground colour
- Background colour
- Text presentation
- Contrast
Step 5: Interpret the measurements
Use relevant experimental evidence, including CamCAST, to understand the potential readability of the measured text.
Step 6: Identify assets that require optimisation
Where the assessment indicates potential readability issues, creative teams can review the design and determine whether changes are appropriate.
This could involve:
- Increasing character size
- Adjusting contrast
- Increasing font weight
- Simplifying the background, removing noise or gradients
- Changing the text treatment, e.g. not using handwritten typefaces.
- Reducing competing visual elements
Step 7: Retest the revised asset
Testing should form part of an iterative process.
The objective is not simply to identify an issue.
It is to provide evidence that can help teams make better design decisions and then assess the revised asset consistently.
What Should Brands Do When an Image Performs Poorly?
A poor glance readability assessment should not automatically mean that an image needs to be redesigned from scratch.
Instead, it should prompt a closer examination of the specific characteristics affecting the text.
For example, if character size is the main issue, increasing the text size may help.
If contrast is the issue, changing the foreground text colour or background colour may improve distinction.
If a photograph is competing with the text, simplifying the area behind the text may provide a clearer visual field.
If several elements are competing for attention, reducing unnecessary information may help the critical message stand out.
The appropriate solution will depend on the design and brand requirements.
The value of objective testing is that it gives creative teams evidence to investigate, rather than simply telling them that an image “doesn’t look clear”.
From Product Image Testing to Digital Shelf Optimisation
Product image testing should not exist in isolation.
Digital commerce teams already evaluate multiple aspects of digital shelf performance, including:
- Search visibility
- Product content
- Product titles
- Pricing
- Ratings and reviews
- Availability
- Images
- Conversion
Image readability adds another dimension to this ecosystem.
A product can be visible without being understandable.
It can attract attention without communicating its critical information.
It can meet brand guidelines while still presenting text that is difficult to read at mobile scale.
Testing helps brands identify this potential gap.
The broader objective is to make product content work harder within the limited time and space available on the digital shelf.
Why Automated Testing Matters at Scale
Manual review becomes increasingly difficult as the number of product assets grows.
A retailer or manufacturer may manage thousands of images across:
- Multiple brands
- Product categories
- Retailers
- Marketplaces
- Countries
- Campaigns
Reviewing every image manually is time-consuming.
It can also introduce inconsistency.
AI-powered assessment allows organisations to apply the same measurement process across large image libraries.
This can help teams:
- Identify potential issues more quickly.
- Compare assets using consistent criteria.
- Prioritise images for review.
- Establish more objective quality-assurance processes.
- Monitor image quality across large catalogues.
The role of automation is therefore not to replace designers.
It is to make objective assessment practical at the scale required by modern digital commerce.
Product Image Testing Should Reflect the Shopper Experience
The most useful image test is not necessarily the one that is easiest for the design team to perform.
It is the one that most closely represents the conditions under which the shopper encounters the content.
That means considering:
Where will the image appear?
How large will it be?
How quickly will the shopper encounter it?
What information needs to be read?
Can that information potentially be read under those conditions?
These questions shift product image testing from a purely creative exercise towards an evidence-based evaluation of digital shelf content.
The Shopper Doesn’t See What Your Designer Sees
Product imagery has become one of the most important forms of communication on the digital shelf.
But the way brands create and review images does not always reflect the way shoppers encounter them.
A large image reviewed on a desktop monitor may become a small thumbnail on a smartphone. Critical text that looks perfectly clear during creative approval can become difficult to distinguish when the image is reduced.
Product image testing helps identify this gap.
By measuring characteristics such as character size, contrast, font weight, foreground and background colour and text presentation, brands can assess potential readability more objectively.
OCR and AI-powered image analysis make it possible to automate these measurements across large image libraries, while experimental evidence from CamCAST provides a basis for interpreting them.
PGI can then provide an estimate of the proportion of UK adults who could potentially read specified handheld text at a glance under the conditions represented by the calculation.
The goal is not to replace creative judgement.
It is to give creative, ecommerce and digital shelf teams better evidence with which to make decisions.
Because on a mobile-first digital shelf, an image that looks readable to the design team is not necessarily an image that has been objectively tested for the way shoppers actually encounter it.
Read next: How Brands Measure Glance Readability: A Complete Guide for Digital Commerce Teams
