How Bad Ads Can Improve Your Performance

george-costanzaThey say mediocrity comes from the perfect implementation of traditional wisdom. So when you’ve tried every single ad optimization tip and trick you know, it might be time for something very different.

Sure it’s hard to stand out in only a little over 90 characters, and even harder in a page full of other ads selling the same thing you are. So when you’ve followed all the best practices to create a good ad but aren’t getting much traction… maybe it’s time to look to George Costanza’s experiment and try the opposite

You’ll probably end up with a “bad” ad and that’s fine. But I don’t mean the type of bad that would ruin your brand’s image. I mean the type of bad that pushes the limits of your advertising guidelines enough to make you stand out.

Here’s a perfect example. Testing these two almost identical PPC ads revealed a 308 percent increase in CPI (Conversions-Per-Impression) when the word “Personalized” was swapped out for “You-nique.”


Now, do keep in mind that Google is generally strict about misspellings, but does allow some leeway for commonly misspelled words and phrases. Here are some examples from the AdWords Grammar and Spelling Guidelines in terms of what is allowed:

  • Common misspellings that the majority of users would understand (like “nite” instead of “night”)
  • Abbreviations that the majority of users would understand (like “info” instead of “information”)
  • Commonly used informal words or phrases (like “lovin it” instead of “loving it”)
  • Misspellings used as puns (like “grate savings on kitchen utensils”)

Whereas you should certainly be bidding on those keywords, I strongly discourage you from including them in your ads, even if you’re allowed to. Otherwise, you’ll end up with ads looking like this:


See how there’s a fine line here? A little unconventional wordplay can have a huge impact on your ad’s performance, but it certainly doesn’t mean you should go butcher your ads and make them bad on purpose. The point here is just to push you past your comfort zone and show you another approach for more ideas for testing.

Does the same apply to other ad channels? Sure it does. Check out the display ad test below from PlentyOfFish’s ad platform:


The result, and I quote: “0.049 percent CTR vs. 0.137 percent CTR in favor of the shit ad in Microsoft Paint.”

Albert Einstein defined insanity as doing the same thing over and over again and expecting different results. So try something different with your ads this week. You’ll never know what kind of results you can get unless you test it. But don’t forget to monitor your performance very closely to make sure you kill off ads that negatively impact your account.

One final but very important tip: come lunchtime today, swap your tuna on toast for chicken salad on rye, untoasted, with a side of potato salad and a cup of tea!

Related reading

Google Ads announce more changes to match types - Challenges and opportunities
five tools to improve ppc ads
How to use PPC data to drive more SEO traffic
Facebook campaign budget optimization how marketers must prepare for September 1, 2019