PPC Marketing Campaign Management

Recently I began a project for an online spice retailer as the lead PPC Campaign Manager.  The company was just launching its online presence and I was starting with a blank slate. 

I decided that this project needed focus above everything else, so we were going to use the One Metric That Matters approach.  We would choose one Key Performance Indicator that matters most, decide on our target goal for that KPI, then move on to our next KPI once we reached that goal.

Step 1: Minimizing CPC

The first KPI I wanted to focus on was Cost Per Click, CPC.  This seems the most obvious choice for a PPC campaign as it would mean the difference between making money on each customer and losing money on each customer.  I set the CPC target at $0.25 because with this CPC and a Conversion Rate of 2% our clicks would be profitable. 

With the help of Google I chose about 60 initial keywords and then I got started writing the first 12 ads.  As expected the initial CPC was high, a little over $2. 

Google’s CPC is determined by 3 main factors:

  • Expected Click-Through Rate
  • Landing page quality
  • Ad relevance

The first step to take was to break up the keywords into tightly-themed groups with no more than 10 keywords per group.  Each ad group would have its own distinct theme with ads that were related to the keywords.  I put the actual keywords in the ad text and created specialized landing pages for each ad group.  The landing page URL also contained the keywords in the ad group.  Adding in negative keywords also helped the ad relevance and CTR tremendously as it kept people that would not be interested in the ad from viewing it. 

By analyzing the organic search results for the keywords I was better able to understand what the consumer was interested in.  Were they searching for how to use the spices they already had, looking for ideas for new spices, or were they looking to buy?  With this information we could customize landing pages for each group that we wanted to target.  We would try to be a relevant brand to them during all of their ‘See, Think, Do’ stages.  (See Avinash Kaushik’s great article on See, Think, Do

After only a couple of weeks of implementing these strategies I was able to get the CPC to below the OMTM target goal.  So I changed the focus from CPC to Conversion Rate.

Step 2: Conversion Rate

Now that I was happy with our average CPC, I felt my time was best used on optimizing our conversion rate.  I was very happy to see that because of the adjustments made in order to minimize CPC, the conversion rate was already above the 2% mark calculated earlier!  I set the OMTM target goal for conversion rate at 4.5% because I believed this was achievable and was the point where my energy would be better utilized focusing on a different KPI.  In addition to using AdWords to optimize conversion rate, I also made many on-site changes.

The website was built on the Magento ecommerce platform so I decided to use Visual Website Optimizer for A/B testing because of the easy-to-implement Magento plugin offered by VWO.  I first used the A/B testing, along with Google AdWords and Google Analytics, to answer some broad questions about consumers.  How much did they care about free shipping, how sensitive to price changes were they, how could I get them to buy more than one item, etc.  After these specific questions were answered, I began doing many tests on the layouts, images, button placement, and so on. 

After about 7 weeks I had the conversion rate to a little over 5%.  I was very happy with this result and it was time to move on to the final step in the project.

Step 3: Continuous Iteration

I was very happy with the progress that had been made, but there is always room for improvement.  At this point I will spend about half of our time incrementally improving CPC, Conversion Rate, and the various other KPIs.  But the other half of the time will be spent trying to make big strides in improvement by trying new things such as a product recommendation engine, new products, a blog, recipe section, and any other ideas that we can put into the ‘Build, Measure, Learn’ loop.