Diagnosing a Drop in PPC Revenue using the Search Data Map

Before coming to Wheelhouse, I worked in a traditional marketing setting that focused more on project management than data analysis. Those of us in the digital marketing field, specifically paid search, know that performance analysis is the brain of our business and drives our decision-making process. So being able to understand and analyze the cause and effect of our paid search performance was key to grasping this business.

This is where David Kennedy, our VP of Digital Advertising, came to share his “Search Data Map”. As you can see from this tree, it provides a framework to answer the question, “Why have my total sales increased or decreased?” in the most simplified form. For this post, I’ll use the example of what to look for if you have a decline in sales, which focuses on the left-most sections of the Search Data Map.

Diagnosing a change in Revenue

When looking for where your decline in sales came from, there are two areas to analyze… To start, it’s helpful to fully understand the variables that combine to provide the metric that has caught our attention. In this case, we know that Total Revenue is a factor of the value of the orders (Average Order Value) and the total number of Conversions, so we start with those two metrics first.

Average Order Value – If you’ve seen a drop in sales check to see if there has been a change in Average Order Value (AOV). AOV looks at the average amount a consumer spends for each order tracked. So, if you’ve seen a drop in AOV, then this would cause your total sales to drop as well.

Conversions – this metric is again the product of other elements of the Search Data Map. In this case, Conversions could decline because of a change in people clicking on our ads (Clicks) or the overall volume of searches for our keyword (Impressions), which we turn to next.

Diagnosing a change in Conversions

If there was a decline in conversions, then there was either a drop in Clicks or Conversion rate, again each a function of other factors within the Search Data Map:

Clicks – When you see a drop in conversions, the first place to check is whether clicks also declined. If so, this change in clicks can come from two places:

• Click Through Rate – Your Click Through Rate (CTR) is the rate at which someone clicks your ad when your ad is being shown. If people are no longer clicking your ad as often as before, this can be because your ad copy is not as compelling for this given keyword or you have a lower average position. This drop in average position can come from a decrease in your bids or if a competitor has come in and outbid you for this keyword.
• Impressions – If your CTR has not changed, see if your impressions have dropped. This drop in impressions can come from a decline in bids or impression share. If your impression share dropped, it could mean you need to either increase bids or budget or improve your ad copy. However, if bids and impression share did not change, then there are not as many people searching for this item or keyword in general. This can happen to more seasonal products like snow shovels or beach towels.

Conversion Rate – If clicks stayed the same but conversions dropped, then we know that our Conversion Rate (CVR) dropped. When looking at why CVR dropped, there are three places to check:

• Landing Page – Is your landing page firing correctly? Does it relate to the keyword and ad copy? If the answer is “no” to either of these questions, then this is the cause of your drop in CVR and your landing page will need to be updated.
• Offer – Did you previously have a special offer or sale running for this keyword/ad group (ie. Black Friday)? If so, then this would have caused a bump in your conversions during that timeframe. I would compare your conversions against a time before this offer ran to see if there is still a difference in CVR.
• Search Query Analysis – The third reason your CVR may be down is from matching to unrelated searches. For example, if you sell leather watches and are matching to searches relating to metal or plastic watches, then those clicks are not going to convert. When this happens, you want to add these search queries as negative keywords so you no longer show your ads for these terms.

Monitoring Cost

Though this does not directly impact sales, it is always important to watch out for your costs. This will ensure you are staying on budget and will also raise any flags in sudden spikes or drops in spend. If you do find a decline in spend, then there is either a drop in clicks or cost per click. Since we already dug into what to look for with clicks, I will focus on cost per clicks (CPC). If clicks stayed the same but spend dropped, then we know that our average CPC dropped.

Cost per Click (CPC) – If clicks stayed the same but spend dropped, then we know that our average CPC dropped. When this happens look at which Ad Groups saw this drop. Once you’ve narrowed down by Ad Group, you can see which specific keywords saw this change. By comparing to a higher performing time-frame, you can see if these keywords had a change in bid. If you did not lower bids, look to see if there has been a shift in where the spend is coming from. It could be that keywords with lower search bids are receiving more traffic than the keywords with higher bids, which would cause you to see a decline in average CPCs for that given Ad Group.

Use the Search Data Map to help diagnose changes in your PPC performance

With any analysis, look for the “what” (ie. sales dropped) and then find the answer to why this happened. Once you understand where to look and how these data points relate to each other you can effectively analyze performance and create positive impact on your account. In our next post on this topic, we’ll move deeper into the Search Data Map, diagnosing underlying factors of changes in Conversion Rate, Clicks and Cost Per Click, among others.

If you’d like to learn more about our approach to managing digital advertising, please get in touch. We’d be happy to help you diagnose your own paid search campaigns using this method.