How does ALT text factor into testing?

Think about how many people open your email before images load. They see your ALT text first. That means ALT text isn't just a fallback for accessibility. It's the first headline a big chunk of your audience actually reads.

So yes, it's absolutely worth testing. Here's how to do it properly.

Set up the test correctly

ALT text should be the only variable between your two versions. Keep the image itself, the subject line, and the rest of the layout identical. If you change the image and the ALT text at the same time, you won't know which one moved the needle. One variable per test, always.

Good patterns to test against each other:

• Descriptive vs. action-oriented. "A photo of our new summer collection" versus "Shop the summer drop. Limited sizes left"
• Benefit-led vs. feature-led. "Save 20% this weekend" versus "New arrivals for August"
• Blank vs. something. If your current setup has empty ALT text, test that against literally any intentional copy. You'll almost always see a difference.
• Styled vs. unstyled. Some email clients (notably Outlook and Gmail) support basic inline styling on ALT text. A bold, colored CTA phrase can behave more like a button when images are blocked.

Measure it the right way

Here's where most people get tripped up. If you measure only total clicks across the whole send, you're mixing two completely different audiences. Recipients who loaded your images are reacting to the image. Recipients who didn't load images are reacting to your ALT text. Combining those results hides what's actually happening.

The right approach is to segment by image-load behavior. Some ESPs (like Klaviyo and Salesforce Marketing Cloud) give you enough event-level data to see which subscribers loaded images and which didn't. If your ESP offers that, use it. If it doesn't, you can still look for patterns by client. Apple Mail with Mail Privacy Protection enabled will pre-load images automatically, which skews your image-load data anyway. Factor that in.

Look at click rate as your primary metric for ALT text tests. Opens are unreliable for this kind of test because Mail Privacy Protection pre-fires open pixels regardless of whether a human actually engaged.

A few things to keep in mind

ALT text for purely decorative images (background textures, spacers, dividers) should be empty. You're only testing ALT text on images that carry meaning or serve as a CTA. Testing a blank spacer image's ALT text is a waste of a test cycle.

Make sure your sample size is large enough before reading results. A few hundred clicks isn't enough to call a winner. If your list is small, run the test over a longer period or against a bigger segment before drawing conclusions.

If you want to go deeper on how image choices affect engagement beyond ALT text, the impact of images on CTR and conversions is worth reading alongside this. And if you're just getting started with A/B testing images in email more broadly, that's a solid place to begin.

Not sure if your ESP gives you the segmentation data you need for this kind of test? Feel free to ping us at our SOS hotline and we'll help you figure out what's possible with your setup.