Digital marketing makes testing almost too easy, but too much testing can be worse than not testing at all. By not testing, you will discover that what you are doing will either work or not work. By over testing, on the other hand, you might never collect enough data to truly know what, if any, of what you are testing is working. But before we dive too deep into things, what does “working” mean in the sense of testing?
While every business needs to focus on their bottom line, not all tests need to declare a winner based on bottom line results. Think of the process that a new client/customer goes through as they interact with your business. What stage of the process do your ads live - do they attract new traffic to your website to start the process, or reach previous visitors to help them move along, or even finish, the process by generating the lead or sale? Whichever stage you are advertising for, make sure your test and its goal match the stage. Here is a quick (and over simplified) chart that matches tests to goals:
The Table above highlights a few aspects of your ads that can (and should) be tested, but what else can be tested? Again, think of the whole process when considering what to test. One way to go about this is to break down the process into 3 stages - 1) pre-ad, 2) the ad copy and 3) post-ad.
This includes everything that happens prior to the user seeing your ad - placement of the ad, targeting settings, keywords (match types), bids, and more.
While the obvious parts are the Headline(s) and description lines, don’t forget about the display URL as well as ad extensions that also show when you ad gets an impression. These extensions can be additional bits of valuable information or alternative CTA’s to better match a searcher’s intent.
You got the Click, Great! Now what? The Landing Page - the user experience (load time, ease of use, clarity, etc.), the copy and images, the CTA, an offer (free trial, download, or other) and so much more. The landing page itself is a gold mine of testing opportunities, but remember don’t over do it. Your tests need to collect data to declare a winner in order to be successful.
Again, the ultimate goal is more business that gives your bottom line a boost. However, not every test should be focused on the bottom line. Set up multiple headlines to see what best attracts attention and gets clicks. Use different CTAs on your landing page(s) to increase leads and conversion rates. Go beyond advertising and test your follow-up procedures on your leads to generate more sales. Just always remember 1) only test a few things (or even just one) at a time, 2) know what KPI metric you are measuring, 3) collect enough data to declare a winner and 4) Move on to the next test.