Are your email results as good as they could be? For most marketers, the answer to that question is no.

Following email best practices is good when you’re getting started, but the only way to optimize your emails for your particular audience is to test what you’re sending. Most commercial email marketing platforms allow you to run A/B tests, where you send two different versions of the same campaign so you can track which one performs better.

While you can try almost anything when A/B testing, the key is that both versions of the campaign must be identical outside of a single element or you won’t be able to identify what factor influenced an increased (or decreased) open or click-through rate.

If you send a lot of emails, you can test different elements quickly, but if you don’t, you’ll want to prioritize elements based on what you’re trying to improve. For example, you can try different calls to action if your open rate is fine but your click-through rate isn’t satisfactory, or focus on subject lines and other inbox cues if both rates are suffering. Here are seven elements you can test with your next campaign.

1. Send times

Everyone wants to know exactly when they should send their emails. Type the phrase “best email send time” into Google and you’ll see more than 1.2 billion results. MailChimp’s Email Genome Project has analyzed millions of emails sent to billions of subscribers and found that weekday mornings are best worldwide, but that insight comes with the disclaimer that it’s “probably wrong” if you know anything about your audience.

Your list’s optimal send time can vary greatly from another’s, and if you segment your list, it could be different for each subset. MailChimp offers a Send Time Optimizer tool that suggests a time based on who’s in your list, but the only way to know for certain is to send a bunch of emails at a bunch of different times and track engagement.

2. Sender name

Your sender name might be the single most important factor influencing open rates. It’s a signal recipients use to determine whether or not your message is worth their time, often without even thinking about it and before they read the subject line.

The name you choose needs to show not only who the message is from but also that it’s trustworthy and useful. Names to test include the name of your hospital, its president or another executive, publication titles, or even combinations like “First name at Hospital name.”

3. Subject line

The subject line needs to tell recipients what’s in your email and convince them it’s valuable enough that they shouldn’t just delete or ignore it. Like blog titles, subject lines hook people and draw them in to read your content and calls to action. Everything from the length of a subject line to personalization factors like a first name or location to individual words can and should be tested.

4. Preheader text

Another inbox cue that’s often forgotten is the preview text or preheader. In many email clients, the first several words in the body of your message appear next to or underneath the subject line, but in many emails, it just says “View this email in your browser” or other generic text.

Testing preheader text can help it complement your subject line and drive additional opens. Like the subject line, it’s ripe for personalization and further calls to action.

5. Content

Do your recipients prefer long emails or short ones? Do they prefer certain types of content, and can you re-order content to increase click-through rates? Can your headlines be more effective? Should you include more images? Fewer? Should they be larger or smaller? The answers to these questions can dramatically influence click-through rates, and the only way to find them for your particular audience is to test each one.

6. Calls to action

Whether it’s “Read more” on an article or “Register” on an event, the call to action communicates to readers what they should do next. It can be a simple text link, a graphic, or a button, and everything from the size to the color to the position can help determine whether readers actually take the action you want them to take.

Graphics can communicate visually and be processed more quickly than text, but they’re also hidden by default in a large percentage of email clients. Email-safe buttons eliminate that concern, but the default design or the one your designer suggested might not be the best one for catching readers’ attention and driving clicks.

7. Plain text

It may seem counterintuitive, but many tests suggest that plain text emails outperform HTML emails in both open rate and click-through rate. There’s nothing to distract readers from your copy and calls to action, and it makes marketing campaigns look less like, well, marketing campaigns. It might even help messages reach your recipients’ primary inboxes in clients like Gmail because they’re not cluttered up with HTML and fancy styling.

Given the headache that getting HTML designs to work in all email clients and on all devices can be, why NOT give plain text a try?