We should all be testing our PPC ads, but sometimes obstacles stop us. Whether it be getting approval, lack of time or some other excuse, there are several ad tests that are small that have made big differences in my accounts. The setup for each is simple and the results are worth the 15 minutes that it takes to use Excel.
There are 35 characters that can be used for display URLs. The only rule for display URLs is that it should have the same domain as the destination URL. While a part of those 35 characters have to be the website, many website domains will not use all of the available space. The important thing to remember is that the display URL should be the only variable when running this type of test:
In the test below, the display URL was the only feature that was changed between the two ads. For the test, each ad ran side by side. One had the city name in the display URL and the other had the actual product that we offer.
Click-thru-rate decreased by 3% for the ads that had the city in the display URL, however, all other metrics saw improvement. As you can see, the display URL with the city name had a 68% better conversion rate and impressions-until-conversions improved by 39%. I like to look at impressions-until-conversion when there is any discrepancy between the winner in click-thru-rate and conversion rate to determine the overall winner.
In order to run a similar test, it takes only a few minutes to download an ad report, change the display URL and upload the new ads to test against the preexisting ads. If putting the product or city name in the display URL does not make sense, there are many other options, such as a call-to-action or an additional feature of the service.
Starting Price vs. Price Range
If I know that putting prices in my ads have proven to be the best performing ads in the past, testing how to word the pricing makes sense as the next step. In the test below, one ad used a starting price and the other used a price range.
The ads that used a selection of prices saw conversion rate increase by 45%. While the 0.11 percentage point change in conversion rate might not seem like a lot, 45% is a significant improvement, especially when dealing with low conversion rate accounts.
There is no Excel function that will change a price range to a starting price. However, find and replace tends to be pretty effective when changing a large quantity of ads.
Sentence case vs. Title Case
In this test, the ads use the exact same words with only the style changing. Title case is when the first letter of all words is capitalized (besides a few exceptions). Sentence case is when the first word and proper nouns are capitalized. When doing this type of test, every word should remain static with only the style being the experimental variable. Here is an example of the two types of ads that were tested:
In the results below, this stylistic change improved click-thru-rate by 12% and conversion rate by 19%.
If you want to run a similar test, PROPER and LOWER functions in excel work effectively to save time rather than changing each description line manually.
Question in Your Headline?
Have you ever tried putting a question in your headline? This ad test is simple to setup and can have an impact on performance metrics. In example below, only the headline changes and the important features remain constant:
In the results from the actual test, a question was posed in the headline and was the single variable of the ad test. Click-thru-rate and conversion rate both saw a 16% improvement to a question being the first interaction with the ad. The combination improved impressions-until-conversion by 40%.
Like display URLs, question in headlines are very easy to setup. Download an ad report, change the headline to a question under 25 characters and upload the new ads.
Cassie recently had an entire case study which dived into DKI and her experience with these types of ads. Because this great post already exists, I won’t waste words on the explanation. My results were similar to Cassie’s:
Site visitors, who clicked the DKI headline were 29% more likely to convert than the other visitors.
Having shared this data, do not assume that what works in my accounts will work in yours. We test ads for a reason: every business is different and the ways of appealing to the right type of visitors do vary.