eCPM: Shooting For The Moon
The Buddha is quoted saying, "My teachings are a finger pointing to the moon. Do not get caught in thinking that the finger is the moon."
Though this was pre-email marketing, this is how we treat eCPM* at Loop7. We love eCPM because it's a snapshot of how "potent" a send is. By dividing the total revenue by how many people it was sent to, it adjusts for size differences across lists. Then we can accurately compare an offer run to a list with 10,000 subscribers vs. one with 200,000. I think most list managers think in these terms, but not necessarily with eCPM as the most important metric. The eCPM is the final say - the moon - and opens, clicks, and EPC (earnings per click) are all just different fingers pointing to it.
All too often, people get caught up on EPC..."hey Affiliate, you had a $2 EPC on that send, so it must have done great for you!" but the reality is, the click rate was terrible, so eCPM did not hit benchmark. Because eCPM translates more directly toward revenue, it is a better measure for overall “success” or “bang for your buck”.
If a send has a low eCPM (translating to low revenue), we can analyze the other metrics like opens and clicks. If it has a low open rate, but seems solid in every other way, we'll test a different subject line. Maybe we just didn’t grab enough subscribers’ attention?
On the other hand, if a send had a high eCPM, we can also figure out why it succeeded. Maybe the EPC is especially high, so we theorize that the body of the email is optimized and integrates well with the landing page. We start with the eCPM, and then run diagnostics as needed.
But offers can "heat up" and “cool down” quickly, and true testing takes time (as I mention in the From Lines post). It's often more efficient and profitable to move onto the next new offer or run something "tried and true" than tinkering until we turn an underperforming swipe into a "winner".
eCPM Benchmarks Across Segments
Establishing different eCPM benchmarks for the different segments within a list allows for more accurate and granular analysis of a send.
Two common segments on a list are a “buyers” segment (subscribers who have actually made a purchase),and a “leads” segment (subscribers who have abandoned cart, signed up for a newsletter, etc). These two segments will not behave in the same way, and need different benchmark eCPM’s that reflect their behavior. A $30 eCPM for an email sent to a “buyer” segment could be low, while a $30 eCPM to the “leads” file is double the benchmark.
Likewise, a segment of subscribers who have opened in the past 3 days won’t behave the same way as a segment of subscribers who haven’t opened in the past 30 days. Or maybe part of your list came from purchasing a low-ticket item, and another segment is from purchasing a high-ticket item. By breaking out these segments, we can establish benchmark eCPM’s relevant to each one. Then we can accurately assess performance and optimize future sends based on those results.
The Final Say
A high eCPM typically means that the other metrics are solid as well. So in order to get the most accurate snapshot of a send’s “success”, we look to the eCPM. That is the deciding factor on whether or not to re-send an offer or try a similar creative. Once each segment has an established eCPM benchmark, we can easily analyze performance of each send. In sum, the eCPM is our way of staying focused on the moon, without getting too distracted by other pointing fingers.
*The “effective cost per mille,” or more simply "effective cost per thousand impressions" is how much revenue a send has generated, divided by how many people it was sent to, multiplied by 1,000.
Sent to: 50,000 subscribers
1,500/50,000 = .03
.03*1,000 = $30 eCPM