Published on January 25th, 2010 | by Formstack4
Creating A Successful PPC Campaign
In the conversation of organic versus paid search, the latter can sometimes receive a bad rap. To the small business owner or novice marketer, paid search can be a scary subject with all the rumors of high cost, low ROI and click fraud. The truth is that paid search can be profitable and provide valuable data for your business if done correctly. You just have to have the right system in place.
Building a successful SEO program is important, but it can take several months before you see any type of impact on your search placement. The benefit of PPC is that it provides immediate, qualified traffic to your website, and you can have a program up and running in a matter of days. This allows you to collect information about your visitors, sales funnel and what search terms people are using to find your site. The information you obtain from PPC can help you with your SEO.
PPC Tools of the Trade
At Formstack, we use both Google AdWords and Google Analytics to track our PPC campaigns, along with some advanced in-house tracking systems. AdWords allows us to track what happens during the initial search and click of a keyword, and Analytics gives us valuable post-click data. These tools have helped us discover a few techniques to increase the efficiency and effectiveness of our PPC campaigns.
Tips and Tricks
Our first technique is that we optimize our landing pages. Many businesses make the mistake of sending all of their paid search traffic to their homepage. Imagine your website is a store with lots of different items for sale. If you only have one entrance, people have to wander around searching for the one item they need. Wouldn’t it be helpful to have several entrances where the customer can walk in and is immediately greeted by the product they were searching for? We create specific landing pages based on our keyword campaigns and direct users to them according to their search criteria. This is a very simple concept, but often overlooked.
Another one of our tricks is to use UTM tags to build very detailed URLs that pass information along to Analytics. This enables you to see exactly where your traffic is coming from. Below is the anatomy of a destination URL using UTM tags.
Test, Test Again…Then Test Some More
A good PPC campaign doesn’t end at tracking. You should constantly test and optimize your ads.
I read an article recently called “Split Testing Adwords: You’re Doing It Wrong,” that got me thinking about our testing strategies. Dan Thies explains that although the way we perform A/B split tests for our PPC ads is normal, it’s also “completely, utterly and totally wrong.” Ouch.
So what is considered “normal” when A/B split testing? Our method has always been to run 3-4 ads at a time with only one variable changed in each ad such as the ad title, display URL or call-to-action. After roughly 1000 impressions, we determine which ad is performing best, pause the underachievers and create new ads to test against the winner.
Thies’ argument against this method is that a winning ad has performance history and is more likely to be given a higher ranking and placement compared to your newly created test ad. Placement is obviously important for increasing impressions, clicks and conversions so you cannot accurately compare ad A and ad B if they aren’t on an equal playing field.
To test this theory, we made three exact duplicates of our control ad (the winning ad), created one test ad (the competitor) with different ad copy and kept our original control ad running. This method spreads out the impressions so your test ad, which could potentially fail, doesn’t receive a large percentage of impressions.
Our plan is to pause the test ad if it’s obviously a failure, or pause the control ads one at a time and keep the test ad running if it’s doing well. Once the test ad has enough performance history and we’ve paused all of our control copies, then we can have a 50/50 matchup vs the control.
To Be Continued…
We’ll let you know how the test goes once we receive enough impressions. In the meantime, we’d love to hear your thoughts on this and other testing methods along with your PPC success stories or failures.