Always Be Testing

Testing ad copy and keywords used to be the secret to PPC success.  These days, it is no longer a secret as it has moved into the “must do” category for any online advertiser.  While we encourage testing in every way, there is something that the experts aren’t telling you.

The most popular advice about keywords might be doing more harm than good.



OK, here’s where we’re coming from.  Ultimately, you should nail the right overall messaging for your product before you start experimenting with your keywords.  Once you have captured the right messaging, you can work toward keyword adjustments, but hold off on fine-tuning terms and adding long-tail keywords until then.

You’ll hear us promoting long-tail keywords, and we’re definitely believers.  But if you’re at a particular stage of your campaign, proceed with caution:

Long tail keywords could backfire on you.  We’ll show you why introducing those too early could sink your efforts.
This order of testing will be your best strategy:

  1. Choose your keywords
  2. Create ad copy variations
  3. Test and optimize your ad copy variations
  4. Test and optimize your keywords

Why the gap between choosing your keywords and testing them?  If you have 50 keywords bringing you a total of 1000 clicks, estimate that each keyword is pulling close to 20 clicks on its own.    Depending on the activity in your campaign, accumulating enough clicks before you see a clear direction can take a long time.
Compare that to testing a few ad variations against each other, and ad copy conversion tests at warp speed.


Ad copy test execution

We found an ad comparison tool that can tighten your ad copy testing even further.  So now testing three separate elements (headline, ad copy lines #1 and #2) happens in fewer steps to get you optimized more quickly. *

This Taguchi-based ad optimizer builds on an effective method that singles out the best performing elements with fewer trials.  You will get to statistical significance twice as fast and for half the cost.

….so, while you are building your traffic for more reliable keyword results, test your ad copy simultaneously.


Resist the long-tail lure

Long-tail keywords-more descriptive and targeted terms than traditional keywords-are incredibly valuable, but they can derail your testing when they are brought in too early.

If building volume is critical to getting accurate feedback about keywords that work for you, then saddling your campaign with long-tail keywords works against you.  Long-tail keywords promise higher conversions, but their low-traffic production would steer you to inaccurate representations of whether or not your ad copy works.

Early in your ad copy testing, avoid the long-tail (exact match) keywords until you have polished your overall message. Keep your short-tail keywords as broad match (or modified broad match) since you can invest more confidence in these base terms. Once you are confident in ad copy that converts for you, introduce the long-tail terms and continue to tweak your ad copy based on activity spread across more solid keywords in the campaign.   You will also be able to start weeding out some of the broad match keywords where you started.

Over time, start with a wide net and whittle down  into a diverse combination of long-tail keywords and short-tail; exact match and broad match taken from the keywords that brought the right results.  It’s not just about sticking with strong performers, but watch for the right opportunity to increase your bids on these, too.

Now that you have the discipline to test properly, you need material.  Follow our tips to Create 100 Different Ad Variations in 5 Minutes.
*SpyFu is not affiliated with in any way.