Text to Image ratios in email

One of the questions I get from folks about delivery is what the optimal text to image ratio there should be in an email. I’ll be honest, I hate this question. Why? Because the question is actually irrelevant. I’ve seen companies with a single image and no text get to the inbox. I’ve seen companies with no images get to the inbox. The text to image ratio is not going to make or break delivery.
Sending mail that the user expects and wants is the crucial part of delivery. If the user wants a single image? That’s the right ratio. If the user wants a readable message with images turned off? That’s the right ratio. Fretting about 20/80 or 18.5/81.5 or 43.256666/56.743334 is getting buried in details and missing the bigger picture.

Emails should be readable. These days, being readable on mobile is critical. This is the single best argument I can think of against one-image emails. And there are still folks who read email with images off by default. Being one of them, I think it gives me insight other delivery folks lack. If I am engaged with a brand, how do I show it outside of loading images? What kinds of things let the brand know I’m still a happy recipient?
If anyone tells you that your delivery problems are the result of a bad text to image ratio, run away. There is zero chance that’s actually true. Filters aren’t going to just look at the ratio and block ratios that fall into a certain range.
What filters are going to do is take the text to image ratio as part of the fingerprint of an email stream. They’re going to recognize certain factors about emails that users like, and factors about emails that users don’t like. Some of these factors will be things like the text to image ratio. But the “wrong” ratio isn’t why mail is being filtered. Mail is being filtered because the recipients aren’t interacting with it enough and the ratio is just one part of the way the ISP identifies it.
Stop fretting over your exact text to image ratios. The right, or wrong, ratio isn’t a true factor in delivery. Instead, focus on creating an email stream that users want, expect, and engage with. This starts with address collection, but collecting accurate addresses isn’t enough. You also need to provide value to them.
Sending mail users want is the key to reach the inbox. Doing that takes time and investment by the sender.

… and bad acquisition practices
ESPC meeting

Related Posts

Images, again

It’s a new year, but an old problem. Email with unloaded images.
Sure, you should be including critical content as text, and/or including alt-text as a normal part of your creative design process, but at the bare minimum you should look at what your mail looks like without images.
The last thing you want to do is send out email with just one strong call to action – the unsubscribe link.

Read More

The source of deliverability problems

Most deliverability problems don’t start where many people think they do. So very often people call looking for deliverability help and tell me all about the things they’re doing to reach the inbox. They’ll tell me about content, they’ll tell me about bounces, they’ll talk about complaints, engagement, opens and clicks. Rarely will they bring up their list source without some prompting on my part.

The reality is, though, that list source is to root of deliverability success and deliverability problems. Where did those addresses come from and what do the people who gave them think you’re going to do with them?
Outsourcing collection to a third party can cause significant issues with delivery. Letting other people collect addresses on your behalf means you lack control over the process. And if you’re paying per address, then there monetary incentive for that company to pad the list with bogus addresses.
Sometimes there are even issues with having your own employees collect addresses from customers. For instance, a retailer requires sales associates collect a minimum percentage of addresses from customers. The company even ties the associates’ evaluations to that percentage. Associates have an incentive to submit addresses from other customers. Or a retailer will offer a discount for an address and customers want the discount but not the mail, so they give a fake address.
All of these things can affect deliverability.
Address collection is the key to delivery, but too many companies just don’t put enough attention to how they’re collecting addresses and entering into the relationship with subscribers. This is OK for a while, and delivery of small lists collected like this can be great. But as lists grow in size, they come under greater scrutiny at the ISPs and what used to work doesn’t anymore.
The first step to diagnosing any delivery problem is to look at the list. All of the things ISP use to measure reputation measure how well you’re collecting addresses. Changing IPs or domains or content doesn’t change the reason mail is being filtered. It just means the filters have to figure out something new to key on.
Want great deliverability? Start with how you’re collecting addresses.
Want to fix deliverability? Start with how you’ve collected addresses, how you’ve stored them and how you’ve maintained them.
 

Read More

Zombies are real but less of a problem

A few years ago I wrote a series of blog posts about zombie email addresses. Zombie addresses are those that someone owned and used and interacted with, but for whatever reason stopped logging into and checking. This series started with the time before the zombies, and moved on to the zombie uprising. Then discussed how they don’t eat brains, but they do love to take a bite out of deliverability. Smart marketers, however, can defeat zombies by the judicious application of the double tap.
portrait of a Zombie computer maniac looking camera from side
Since that series of blog posts a few things have changed. The biggest thing is that the webmail providers are being much more aggressive about disabling email reception at addresses where folks don’t log in. I have a few addresses on different providers I use for testing purposes. I have to remember, though, that I need to log into them before sending test messages. If I don’t, they generally bounce.
This doesn’t completely remove the challenge of zombie addresses but it does make it easier for regular senders to purge their lists of zombies just through their normal bounce handling. No double-taps needed.

Read More