EPC’s are one of the worst metrics to use for optimizing a campaign. They are extremely misleading and will hurt campaign’s performance. Yes, even the data you are collecting on a 3rd party tracker will trick you. EPCs are calculated by taking the total offer page clicks and dividing them by the total revenue generated. Even a minimal amount of clicks being dropped can skew the EPC’s into someone else’s favor.
If you are not collecting data, they can also be easily manipulated by the offer owner and sometimes the network’s. Some networks won’t only count the 1st original ip, and will count the same (visitor) click when they return. If split testing networks, this can artificially inflate a offers EPC. So make sure you keep that in mind if you are only using the networks data to track. Using ROI as the base number is the safest bet. Its always constant across every test you make, and all the data.
Let’s say you are split testing two different offers on spend, and offer #2 is more popular. People are more aware of it and trusting of the brand. The payout is lower, but the branding gets you a higher landing page ctr, and cr%. In the example test, assume we gave each offer the equal amount of spent in traffic.
Offer Split Test Example:
Offer 1: Payout $39 | 1000 / $1,560 (40 leads ) = 1.56 epcs.
Offer 2: Payout $32 1030 / $1,600 (50 leads ) = 1.55 epcs
It would only take 30 clicks to make the offer that generates you less money to look like the best one. This is why you want to use ROI as the metric to optimize campaigns.
More so when using automatized optimization, because of similar reasons. The system will focus on a funnel that generates the highest EPC’s but at the end of the day. The other offers or funnels might be the ones that make you the most money per dollar spent. You want to make as much money as you can right?
Auto Optimization Example:
Funnel #1: 750 / $520 (65 leads) = 0.69 epcs
Funnel #2: 700 / $496 (62 leads) = 0.70 epc
Funnel #1 drove more clicks to the offer pages, and its EPC’s were lower. You spent the same amount of money per funnel on the test. It made more money than the second funnel due to the extra clicks sent.
This is why you want to use ROI. At the end of the day, its about how much money you get back per dollar spent. Don’t get lost in the over all Big picture and look into the fine details of the campaign.