How do you ensure the viability of your headlines? Test and test again. Here are five ways.
Writing a great headline takes work. It’s not an exact science because of varying topics, audiences, length requirements, writing styles and so on. However, one thing you can keep consistent is testing your headlines. Headline testing gives you the highest chance of getting those golden clicks. How do you ensure your headlines are the best they can be? In this post, we’ll post explores five ways to test your headlines and look at a few solid examples.
A good rule of thumb is to come up with three or four headlines for an article or a webpage and then test those headlines against one another. It’s not necessary that you use more than one of the testing techniques described below each and every time, but if you want the exposure, it’s worth the effort.
1. Test your headline on Twitter.
The testing strategy of pitting two headlines against each other was made popular by Buffer. This method is fast and flexible and after testing it out myself, I found that the method works well. Here’s the process:
- Choose a testing time that will give you a lot of feedback. When testing headlines on Twitter, you want to make sure you’re testing at a time when your audience is on Twitter and interacting with your account specifically. This generally means during the week and in the afternoon, but you can use your Twitter metrics to see when most of your followers are online.
- Tweet both of your headlines an hour apart. If you find that you have a lot of activity in the morning and then a lot at night, still be sure to tweet an hour apart as opposed to at completely different times of the days. This will help you eliminate the variables when testing both headlines. Below are two headlines that I tested:
- Use Buffer or another tool to look at the performance data. After 24 hours, analyze the data using a tool like Buffer. The data will tell you which tweet got more favorites, retweets, mentions, clicks and potential. Below are the results I saw for my two tweets above when I looked at the Buffer analytics tab:
As you can see the metrics are a bit split, which brings me to my next point.
- Understand what metrics matter most for you. In general, potential is the least important. This will help let you know the value of those tweeting your articles (essentially, if they have a lot of followers), but nothing has actually happened. For this reason, you should base your decision on the other metrics.
Once you start testing your headlines this way or testing through other mediums, you’ll start to see a trend in your metrics. This will help let you know what metrics are most important to you. If you’re new to the testing game, just stick with number of clicks—this is the most meaningful for most of us.
Once you’ve found your “winner,” you don’t have to stop there. Consider testing another variation of the headline that came out on top. If you want to be extra thorough, test two headlines again. For example, because the second tweet earned more clicks (but still had engagement), which is most important to me, I could test variations of that headline such as “How to Help Your Holiday SEO” or “How to Succeed in SEO Over the Holidays.”
2. Use promoted tweets for paid testing.
Another strategy is to use paid testing through Twitter to test headlines. This is ideal is you don’t have a large following and therefore can’t generate a good amount of data. By using promoted tweets, you can try your headings out on a larger audience. Below is an example of what a promoted tweet looks like to Twitter users:
To be fair, it doesn’t make sense to do this every single time you want to test a headline because you do have to pay for it. Nonetheless, if you have a special offer, ebook or something that could bring you money in addition to more clicks and visibility, it could be worth it occasionally. You can learn more about getting started with promoted tweets here.
3. Use email for A/B testing.
This way of testing is, again, similar to how you would test on Twitter. Choose two headlines and test them against one another by setting up a subject line test for the email that you send out to your subscribers. I learned about this technique from the tool Campaign Monitor and found it to be another great way to test headlines. The process is as follows:
- Select “subject line test” from the testing options. Here is where you will include your two headlines. Then, write your email message as normal.
- The tool will then send one version of your email to part of your list and the other version to the other part and measure results.
- Results are based on the best performing email, which means the emails that got the most clicks. The “winning” email will then be sent to the remainder of your email list.
Below is an example from the tool that shows what a report using this system looks like. This data will be sent straight to your email:
The cool thing about testing your headlines with this tool as opposed to a platform like Twitter is that the sample size is controlled. The same number of people will see the two headlines.
4. Use the “25 headlines” Upworthy approach.
Upworthy is notorious for their headlines as well as for their process for creating headlines. To put it bluntly, they test rigorously. They employ what they call a 25 headline rule where you have to come up with 25 headlines for every single article you create. This helps you think outside the box, but not overthink each headline. According to Upworthy, they wrote a post about intelligent monkeys, and one headline garnered 59 times more viewers than the other.
So how exactly they do to actually test? They have a custom click-testing system they call the “magic unicorn box.” Because this system is used internally, it’s not publicly accessible. However, that doesn’t mean you can’t use their strategy of writing a lot of headlines quickly and then testing through other methods (Upworthy has mentioned that they test headlines on Twitter as well). In general, they look at clicks per share and shares per view. Also keep in mind that A/B testing can affect the search engine results on your post, so make sure you’re doing things correctly on the SEO front.
Of course, 25 headlines for every article may not be realistic (or maybe it would work for you, in which case go for it!). Still, the testing mentality can apply even if you have less time or less to work with.
5. Use the CoSchedule Headline Analyzer.
If you have limited time and want to get feedback on the viability of a headline quickly, CoSchedule recently released a Headline Analyzer tool. Simply type your headline into the form field, click “Analyze Now” and get a score for your headline. The tool gives you tips on how to increase your score and keeps track of previous headlines you’ve tested.
Examples of Great Headlines
Sometimes in order to know what to test, it helps to see examples of different headlines and the approaches the headlines used to find success. Below are some of my favorite headlines and reasons why they work so well:
BoostBlogTraffic uses emotionally charged headlines.
Upworthy uses storytelling with a little bit of mystery.
And Buzzfeed likes numbered lists (and uses subheadings, discussed in the next section).
If the examples above aren’t bringing you any inspiration, check out this article from Unbounce for a general checklist of what makes a good headline. This will give you a starting place and give you material to test that is actually worth testing.
A Word on Subheadings
You may notice that many publications and blogs use a heading and then a subheading, or tagline, just below the title. This can be a great way to add something to your headline and something you may want to try testing. If both of your headlines worked great in your initial tests, see if you change one into a subheading. Subheadings don’t usually help you in organic search, but if someone is scrolling through your website a tagline could be the thing that catches his/her eye. Think of it just as a little extra ammunition.
Here are examples from the iAcquire blog, which uses taglines to tell what the article covers.
Do you have your own method of testing headlines? Do you have any examples of great headlines you’ve seen in the past or on any particular website? Let us know in the comments below.