One of the best aspects of managing a pay-per-click campaign is the completeness of data. It is the only advertising medium that can give you a fairly accurate representation of how many eyeballs see your ad, and it’s the only advertising medium that can regularly associate a user behavior with an advertising placement.
But with all of the data that’s available to an account manager, how can you determine what to look at, and which metrics truly matter? Here’s a simple guide:
Click-through Rate (CTR)
How to Calculate It: Clicks divided by impressions
What It Tells You: What percentage of the time an ad display leads a user to your site; which types of ads are most relevant and most stand out to searchers on a SERP
What It Doesn’t Tell You: Whether your ad actually left an impression (no pun intended, I swear) on the user who clicked or did not click; whether your clicks were the result of well-crafted ad language, or simply due to aggressive bidding
Best Used For Evaluating: Quality of your ad text
Who It’s Most Useful For: A new or re-focused website focused on bringing in new traffic; Sites primarily interested in exposure
Cost Per Acquisition (CPA)
How to Calculate It: Total spend divided by total conversions
What It Tells You: Your average cost for each conversion; how much you are spending for each sale, lead, or site registration
What It Doesn’t Tell You: Depending on the conversion, whether the acquisitions are actually improving or can be expected to improve your bottom line; whether you’re bringing in new customers, or simply generating repeat conversions from site loyalists
Best Used For Evaluating: Effectiveness of your bidding strategies
Who It’s Most Useful For: Site owners operating on a limited budget, seeking to exploit only the most efficient of keywords
Conversion Rate(CVR)
How to Calculate It: Total conversions divided by total clicks
What It Tells You: The frequency with which ad clicks are leading to a desired user action
What It Doesn’t Tell You: Whether mid-to-low ad position is effectively leaving your site under-exposed
Best Used For Evaluating: Quality of your selected landing pages
Who It’s Most Useful For: Business-to-business sites seeking to bring in a large number of prospective clients at the start of a long sales funnel
Average Sale (Avg. Sale?)
How To Calculate It: Revenue divided by total sales
What It Tells You: Which keywords are bringing in big-ticket purchases; which ad language is most conducive to bringing in high-spending customers
What It Doesn’t Tell You: Whether the high-ticket sales correspond to high profit-margin items; how much you’re spending for these high-revenue purchases
Best Used For Evaluating: Whether or not your ad text speaks to the potential large-revenue consumer
Who it’s Most Useful For: Experienced e-commerce site owners seeking to tweak ad language or landing page selection to drive campaigns to peak performance
Return on Investment (ROI)
How To Calculate It: Net revenue divided by net advertising spend
What It Tells You: How much benefit you are getting for the money you are spending; in theory, all you need to know
What It Doesn’t Tell You: Whether this ratio is the best way to evaluate your campaign’s success, or whether you are over-focusing on a ratio rather than on the bottom line
Best Used For Evaluating: The interaction between all elements of your online campaign
Who It’s Most Useful For: When calculated properly (taking into account every aspect of campaign revenue and potential revenue), ROI is relevant to all website owners
There’s no perfect metric with which you can analyze campaign success. As this list indicates, the most important metrics really can vary on a client-by-client, or an industry-by-industry basis. In addition, it is important to realize that all of your media campaigns – paid search, direct mail, outdoor, radio, social networking – interact with each other and, as a result, you can never be sure what keyword or ad language truly drove which purchase.
The nice thing about pay-per-click marketing, though, is that you have an idea. That’s a start.
