They say a picture is worth a thousand words. But which picture? Here at Alchemy Worx we wanted to test the performance of images that contribute to a recipient’s understanding of the content of an email. Specifically, we wanted to know what works best when sourcing imagery to accompany editorial content – the pretty or the pretty informative?

What works best as a guideline for sourcing images for your email newsletter content: “pretty” or “pretty informative”?

To answer the question of what kind of image works best to illustrate Email-Worx content, we tested two different images against each other in two editions of Email-Worx:

Test edition 1 – images portraying different meanings of the topic

Test edition 2 – different images portraying the same meaning

Here’s what we found…

1 – Images with different meanings

The subject of this email that the image had to illustrate was A/B Split testing. So we went for one image that made reference to the “A/B” element – a pretty obvious image of bricks with the letters on, and one image that illustrated the idea of “split” – a rather more sophisticated image showing a group of people separated by a fissure in the ground.

2 – Images with similar meaning

This message was about lifecycle email campaigns, so we chose images that both represented a lifecycle but in different ways: bud-to-flower and chrysalis-to-butterfly.  

Analysing the results

Although in all test variants the copy and subject line was unchanged, we still saw different open rates, which makes the click-to-open rate the most appropriate metric to measure the impact of the image.

Why? The total and unique click rates are affected by the open rate, as clickers must have opened the message to be able to click. As images are not visible unless an email has been opened, the click to open rate equalises the tests by removing the impact of the different open rates. The click to open rate represents the proportion of openers who then went on to click – as the copy and format did not change, this is most likely to be a direct influence of the recipient viewing the image.

Results for 1 – Images with different meaning

Open % Total Click % Unique Click % CTOA/B

Bricks image 32% 14.7% 12% 38%

Split ground image 35% 12.1% 10% 28%%

Change 34%

The A/B image had a markedly more successful click-to-open rate. Our hypothesis is that the A/B image is much more closely aligned with the predominant mental image or phrase that people associate with the subject. In short, people think “A/B” before they think “split”, and the bricks image responds very directly to that mental keyword. In addition the placing of 2 similar objects side by side provides an obvious, if loose, visual analogy with the comparing process involved in variant testing.

The split image, despite being more interesting and of better quality, was less successful for the same reason highlighted above. In addition, the image perhaps also introduces other associations – earthquake, rupture, division etc – that are less welcome or relevant to the content, and so cause some semantic interference to the recipient trying to get a quick understanding of what the article is about.

Results for 2 – Images with similar meaning

Open % Total Click % Unique Click % CTO

Butterfly life cycle image 29% 12.3% 9% 31%

Flower life cycle image 34% 12.2% 10% 28%%

Change 12%

The results for the second mailing bore out our hypothesis that we would see only a small difference as both mailings had the same meaning. Both images are good visual illustrations of the key concept in the content, and their comparative prettiness or visual appeal appears to make little difference.

Conclusion: pick images that convey the meaning of your subject matter

These results show the importance, particularly in more editorial contexts, of choosing an image that directly supports the meaning of your message. Images are very subjective, of course, but even when a very “pretty” or “more visual” image was compared to a “less pretty” or “less visual” image, the results were relatively similar because both were delivering effectively the same meaning.

In a content-rich environment like Email-Worx, well-chosen images that act as a visual cue to the message’s contents are more likely to generate clicks than those that are more obscure in their meaning. Images are information: they enable you to convey a huge amount of meaning in a short space and time, and so careful image selection is a crucial element of email performance.

The guideline to choose images that represent conceptually the information you’re trying to convey can be applied to product-focused content too. It suggests that images should illustrate not simply the product itself but an attractive scenario showing the product in action. So for an underwater camera, the best image might not be a simple product shot but rather an underwater scene showing a scuba diver taking pictures of a colourful reef. That kind of image is effective because it contains a huge amount of relevant information. A thousand words indeed.

Last updated: May 12, 2015  Alchemy Worx

‍