Skip to main content

Testing Headline Combinations in Google Ads Instead Of On Landing Pages

Here's what I'm doing to test over 72 headline combinations using Google Display Ads
math for different domains

I heard two different things about headlines:

  1. long headlines perform better than short headlines
  2. keep your headlines to 7-10 words (copywriters)

My plan is to create a super-long headline on a landing page and run multiple ad combinations at it. Whatever gets clicked on in the ads stays in the headline. Whatever doesn't perform well, gets cut out of both the ads and the headlines. Easy ... right?

I'll be testing 4 different variables:

  • Promise/Activity (Verb)
  • Segment (Adjective)
  • Objective (Noun)
  • Brand (mine)

1. Promise/Activity (Verb)

  • Reach
  • Achieve
  • Exceed

These are both different types of performance and different end states. Reaching falls short of achieving, and exceeding outdistances both. We'll see how far this gets us (ha!)

2. Segment (Adjective)

  • Personal
  • Professional
  • Business

I may nix personal goals altogether, but I want to see what gets clicked at what rate first. I need to know whether it's ambitious, career-oriented, or entrepreneurs that like the message the best.

Before we hit 3 and 4, let's check out a few combos of 1 and 2 so far.

  • Reach Personal ...
  • Reach Professional ...
  • Reach Business ...
  • Exceed Business ...
  • Achieve Professional ...

See? Now you're getting it. There are 6 million ways to headline a product, choose 4 million!

3. Objective (Noun)

  • Goals
  • Dreams

Hey, I'm not sure 'Reach Personal Dreams' is a winner, but 'Exceed Business Goals' sounds cool. Whatever the case, the data will let me know.

4. Brand (mine)

  • Imagine
  • Create
  • Execute

The name of my product is the Imagine Create Execute Workbook. It's not sexy, but I like it. It describes the 3 stages you need to go through to achieve any goal. I'm not done, but meanwhile, learn more about this wonderful product/course here:

Before I plaster the headline here, let's talk about ads. Google only lets you put 30 characters in an ad. So, to test this, I have to create a maddening amount of ad variations:

  • Achieve Your Goals
  • Imagine Your Goals
  • Reach Your Goals
  • Create Your Goals

But WHYYY do this with ads? Google Ads rates you with its 'ad score'. It wants the words in your ads to be in the headline on your landing page. You can create lots of slightly different landing pages with slightly different titles, or you can do this.

So, Instead of trying to learn and use google optimize to pair up my ads with my hundreds of headline combinations, I'm going to see what people click on. Do Business people like 'Dreams' or 'Goals' ... do Professional clickers like 'Achieve' or 'Exceed' etc?

So, the super long headline is this:

Reach, Achieve and Exceed Your Personal, Professional and Business Goals and Dreams By Practicing These 9 Steps In The Imagine Create Execute Workbook

Permutation math says that's 3x3x2x3=72 different combinations I'm testing!

And I'm a rookie at this! I'm not event testing things like narrative, positioning, and emotional leanings! Anyhow, its getting light out and I need to do something productive. I figured I'd share this for the other #buildinpublic tweeps burning the early morning oil.

Of course, follow me and hopefully in a few weeks I'll have some data to share on this ... oh I forgot ...

I'm creating Google Display Ads. I have a few pictures and materials, but I'm not testing those. What I am doing is creating an ad set in an ad group and copying each, then changing the 1 or 2 words from the previous set. Why not search?

Search ads are expensive compared to display ads. I get way more information about words even though there's less intent. But I have done a good bit of targeting and excluding to get better data.

Targeting Segments Geographically:

I chopped up my locations into metropolitan areas ... so instead of Philadelphia, it's the Philadelphia dmca neilsen (or whatever) region. I've done that for most of the US, this will tell me what regions like what lingo.

Excluding Segments:

These aren't being shown on beauty, auto, and leisure websites. I categorically excluded a lot of games and mobile apps to make sure I'm reading people at the right place and time. High powered segmenting here!

That's it for now. We'll see how this goes in a few weeks after a couple hundred thousand impressions. If you like this, follow me on twitter here:


When you break down 'Products' and 'Services' into smaller groups, you'll find more ways and examples for how to market what you offer.
Tips are nice, but procedures rule the roost. Here's what to do before you talk to anyone else.
What Other Questions Do You Ask After You Start With Why?