After you’ve been working on a pay per click (PPC) campaign for a while, you tend to get comfortable with what you see in reports. Eventually the numbers from the campaign become predictable and your costs, revenues, and ROI level out. At this point, it’s easy to become complacent, and focus on new campaigns. What you might not realize is that there are hidden opportunities relating to your customers in the numbers.
This article is about helping you squeeze new profits out of old PPC campaigns.
Examine the Right Data
Before you can properly analyze a campaign, you need to identify the most important data points to gain insights at a granular level.
When gathering data for search costs, use Keyword Detail Reports to help you understand the underlying performance of each ad group and campaign, rather than Ad Group or Campaign Reports, which only provide high level summaries. If you only look at the ad group performance, you might not see that a single keyword is driving 90% of the profits, or that a specific ad is driving all the clicks.
Similarly, when gathering data for display campaigns, use Placement reports, rather than Campaign or Category Reports. Placement Reports will show you the specific URLs or sections of a website driving traffic. For example, you might find that a particular placement performs poorly because the prices on the page are better than the prices listed in your ad.
When you’re finished collecting your ad spend data, it’s time to dive into your revenue and profit reports. Just like the previous example, it’s important that you focus on reports that provide the greatest level of detail. Use Transaction Detail Reports that provide you with a timestamp, items purchased, and other important details.
Analyzing these report details on a daily or weekly basis doesn’t make a lot of sense. Campaign spikes and other irregularities can throw you off. Collecting and examining data over the duration of a month, a quarter, or even a year, will help normalize the data, and provide more reliable insights.
Use the Data to Understand the Customer Experience
Once you’ve gathered all of the data, you need to make sense of it. For this step, you’ll need to open up Excel, and pour yourself a fresh cup of coffee.
Start by aggregating all of the data you collected into two spreadsheets. I typically put all of my cost data into one spreadsheet, and all of my revenue and profit reports into another spreadsheet. If you’re managing high volume campaigns, you might be working with hundreds of thousands, or possibly millions of rows.
After you’ve aggregated all of the data, it’s time to crunch the raw data and get some meaningful information out of it. Here’s the information I like to extract (and why):
1. Search Advertising: Highest/Lowest cost per conversion – It’s important to understand your highest and lowest value keywords, so that you can add more variations of the high converting keywords to your campaign, and eliminate wasteful spending on the low converting keywords. Identifying the best converting keywords not only gives your campaign a performance boost, but gives you valuable insights about visitor intent, and what content you need to provide in order to convert clicks into customers.
2. Display Advertising: Best Converting Placements – Display campaigns are difficult to manage, especially when you don’t know what types of sites are sending the best traffic. Some advertisers can blow a month’s budget in a single day. Reviewing Placement Reports will help you identify the sites that send high quality traffic, and remove those sending you poor traffic. Once you have found the sites that convert, you can use tools like Google Ad Planner to find similar sites to advertise on.
3. Sales: Best Days of the Week and Times of Day – Most advertisers use the same budget and ads over the course of the week. However, if you don’t know what days of the week your prospects are most responsive, you might be wasting a lot of ad dollars.
For example, we recently discovered that our best sales days for our HP Deal Page are from Sunday to Tuesday. The sales volume dropped off considerably from Wednesday through Saturday. This information corresponded with the release of their new coupons. We determined that customers responded to the fresh deals, but by the end of the week, visitors suffered from offer fatigue, as they anticipated new deals coming out every Sunday, so we adjusted our ad spend accordingly.
4. Sales: Best Days of the Year – If you’re doing analysis over the course of a year, you’ll start to uncover a lot of surprising trends in your data. For example, many online marketers assume that Cyber Week is the busiest time of year for online retailers. This may be true for most, but we discovered that we had better conversion rates and higher sales over the 4th of July week for several of our merchant partners. Knowing the peak periods for each merchant throughout the year, will help you prepare ads and campaigns for the best windows of opportunity.
5. Sales: Top Selling Products (at Any Given Point) – Knowing what the top selling products are throughout the year will help you present the most attractive offers on your site at all times. For example, the outdoor clothing company Moosejaw.com features different products on their homepage and their ads to maximize sales for the four seasons. Ask yourself, what is the peak seasonality for your products or services? What are people searching for throughout the year? The data, not your intuition, will reveal your answers!
Next time you reach a plateau with a campaign, and feel like there’s nothing more that can be done, dive back into the numbers to find more hidden profits. The dynamic nature of online advertising and your marketplace mean that these opportunities are constantly changing.
Jeremy Palmer is the Chief Marketing Officer at CouponPal. Jeremy is a vocal supporter and strong advocate of the performance marketing industry. He’s been a featured speaker and trainer at dozens of internet marketing conferences. Jeremy’s e-books, webinars, and videos have been consumed by thousands of affiliates.