How a simple tool helps me write headlines with confidence

In many cases, the success of a blog post or sales page rests on the strength of the headline. Says David Ogilvy, among the most successful copywriters of all time:

On the average, five times as many people read the headline as read the body copy. When you have written your headline, you have spent eighty cents out of your dollar. – David Oglivy

That’s a lot of pressure. It seems like every time I sit down to write a headline, that quote springs to mind.

The next thing that pops into my head is usually an anecdote from some random blog post about someone who changed a comma in a headline and increasing their conversion rate by 10,000 percent…

I’m nowhere near experienced enough at any of this to know what might make a difference.

At best I’m just making an educated guess.

But what if I guess wrong?

It’s paralyzing.

But you don’t have to guess

Lately, when I really need a strong headline for a major blog post or sales page, I’ve been reaching for a simple tool. This tool takes the guesswork out of headline selection and allows me to relax and let the creative juices flow.

What is this tool?

Twitter ads.

I use Twitter ads to run cheap tests that allow me to find headlines that resonate with my audience.

If you think about it, Twitter is a perfect platform for trying out headlines. A tweet is basically a headline–you have just a few words to grab attention and get someone to take action.

Here’s how I used Twitter ads to try out a new headline idea for my self-published book, Sublime Productivity.

Sublime’s secret identity!

This particular test started when I stumbled across an epic tweet by a Sublime Text fan:

The budding copywriter in me lit up when I saw that. Pure gold! Would the “superhero” theme make a good hook for my sales page? I thought it would.

Before investing time actually writing a sales page to try out, I decided to test it out with a Twitter ads campaign.

I originally got this idea from Tim Ferriss in The 4-Hour Workweek. He recommends using Google Adwords to test out product ideas before you invest any time in creating them.

But when Derick Bailey introduced me to Twitter ads, it hit me that they were even better for this type of test. For one thing, it’s cheaper. I’ve yet to have Adwords campaigns cost less than $1 to $2 per click, although I’ve heard it’s possible.

Second, Adwords ads impose a lot of restrictions on the length of your headline and body copy, and these requirements make it impossible to test a headline like you’d typically use on a blog post or sales page.

So, what to test?

To do this, I created a new “promoted tweets” campaign on Twitter. A promoted tweets campaign allows you to compose tweets and pay for placement near the top of Twitter users’ timelines.

You’ll want to test several headlines simultenously. I created one tweet based on the “superhero” quote. I also created several additional headlines that target pain points my audience feels, and I threw in a couple that just sprang to mind as I was setting up the campaign. (These aren’t awesome headlines–I’m a little embarrassed to show my homework here.)

Screen Shot 2014-04-27 at 9.20.13 PM

I used a link to my existing sales page for these tweets. I wasn’t expecting to sell any books this way–it’s really tough to convert “cold” traffic from ads like these. But I needed to send it somewhere, so my book sales page was as good a place as any.

I ended up with a total of 10 tweets that covered a range of ideas. (This is too many, really. In the future I’ll stick with around five. If you have too many tweets, Twitter will usually end up favoring a few and the others don’t get a fair test.)

Who should I show the ads to?

One of the reasons I love Twitter is the way it allows you to pinpoint your audience by looking at who people are following. I built my @SublimeTxtTips account originally by following people who were following the official Sublime Text account. Can’t get much more targeted than that.

I focus my ad campaigns the same way–I target the people following big Twitter handles in my niche. For this test, I started with the @sublimehq and @sublimepackages Twitter accounts, then added a number of well-known web developers and web technologies. These people have raised their hands and expressed interest in my topic, rather than just mentioning the topics in conversation.

I wound up targeting 28 different Twitter handles, which gave me a potential audience of millions of people:

Screen Shot 2014-04-27 at 9.21.23 PM

Once my tweets and targeting were ready to go, I set a budget of $10 a day and let the campaign run for a bit.

How do I gauge the results?

Twitter tracks the number of “engagements” your ads receive–that’s the total number of clicks you get, including clicks on your link, retweets, replies and follows. (Side rant: They also charge you for clicks to expand the chrome that surrounds the tweet, which is a total ripoff and makes me really mad.)

This engagement rate gives you a way to measure the amount of attention each headline attracted relative to the others.

Generally speaking, an engagement rate of 2% and above indicates solid interest. It’s quite possible to hit 3%, 4% or even 5% or better, but landing north of 2% is a decent result.

So how did my “superhero” idea work out? (Click the image for a full-size version.)

Screen Shot 2014-04-27 at 9.20.40 PM

Abysmal! With an engagement rate of 0.69%, it was one of the worst performers.

In hindsight, that’s not really surprising. It’s usually better to be clear than clever in your ad copy, and I was breaking that rule here.

What shocked me, though, was the top performing headline:

Make Sublime Text work for you

That was one I threw in as an afterthought. It doesn’t seem particularly interesting or attention-grabbing, but it achieved a respectable 2.63% engagement rate.

This lead me to later try the same wording as a subject line for an email to my mailing list, and that test mailing received an excellent open rate of 56%.

After that double confirmation, I’m planning to rework my sales page to lead with this headline.

Yes, this approach costs money

I can’t afford to test every single headline I write this way, but when I’m trying to select a strong headline for a book sales page, or picking a title for that epic 5,000 word blog post that I want to go viral?

In those cases, it’s well worth dropping a few bucks on Twitter ads to make sure I pick a winner.

Mike Godfrey - February 6, 2015

Hi Josh,

Seems like a great idea – I’ve just finished my longest blog post yet (only 1250 words, but good for me), and I feel like having some marketing fun, so $10 a day for a couple of days is OK.

My questions:
1. Any ideas where to direct the traffic to, considering the post isn’t published and I don’t have landing pages selling stuff? Just the root of my website?
2. Will my current followers be flooded with tweets for the same content? Do I risk them unfollowing? Does this even matter?

Been following you for a little while and it looks like you’re slowly changing career paths?

Cheers and keep up the good work,
Mike Godfrey.

    Josh Earl - February 6, 2015

    Hey Mike!

    Good questions. Yeah, I think just sending the traffic to your home page is fine.

    You can choose whether or not to include your current followers in your audience – I usually don’t include them.

    And yeah, as you can tell, I’ve been shifting my focus to marketing/copywriting these days. Loving it! 🙂

      Mike Godfrey - February 9, 2015

      Hey Josh,

      So I just launched my first Twitter campaign. My steps for your future readers:

      Tweeted each potential headline with Buffer: a. “#Screenshot #testing with #protractor” b. “How to #test #angularjs #svg directives”

      Created a Twitter promoted tweets campaign targeting: 2 locations, UK and US. Followers of @protractortest and @angularjs. Users like my followers. Keywords as the hashtags in each tweet (6 in total).

      Set a max daily budget of £10 and a max total budget of £10

      Pressed launch!

      I’ll post back with the results if you’re interested?


        Josh Earl - February 9, 2015

        Totally–love to hear how it goes!

          Mike Godfrey - February 16, 2015

          Hey Josh,

          Somewhat delayed, but…

          Headline a got 7,448 impressions with 41 tweet engagements, 0.55% engagement rate.

          Headline b got 4,907 impressions with 20 tweet engagements, 0.41% engagement rate.

          I shared the tweets with Buffer so have link analytics too…

          Headline a got 2 clicks.

          Headline b got 6 clicks.

          So, what gives? Well I think I conducted two experiments in the same test. By using hashtags in my headline tweets I tested both the hashtag audience (#testing had the biggest) and headline engagement. I’m no expert but the results appear flawed.

          Your technique of targeting the big people in your target niche would appear to solve this problem – oh well, next time perhaps…


          Josh Earl - February 16, 2015

          Hey Mike,

          Ah, interesting result. You’re right about the hashtag audiences.

          Also, hashtags tend to get clicked a lot by themselves, so including them will skew your results. You were getting clicks on both hashtags and the blog post itself, which isn’t quite what you wanted to test.

          I’d declare headline b the winner here, even though the other headline got more engagements.

Comments are closed