Early Evaluation of Native Campaigns (5)
Hello all. This is my first post on STM. I'm usually a forum lurker, but I have my own question now. After a two-year hiatus from attempting to make money at this, I have jumped head first into natives! Thank you for taking the time to look at this post.
In order to evaluate campaigns early on, I've been looking at my lander EPCs (revenue/clicks from lander). The idea is to look at the strength of an offer early on an see if I should be investing in sending traffic there.
In the case where I have copied campaigns and offers, using Anstrex, I have been killing campaigns early on. My rationalle is that if my EPCs off of my lander aren't realistic (for example a .05 EPC where a reasonable bid for a click is .15), I kill the campaign before spending 4x or 5x of the payout.
I know that the offer is probably the most important aspect of the campaign, and testing multiple offers could lead to better results. Also, what I'm referring to here is primarily non-English campaigns where there isn't a great variety of offers available.
I realize that ad copy and lander copy contributes to conversions, but if the EPC from a lander is only half or a third of the cost of the click to the lander, I wonder if it's not a good idea to kill the campaign sooner rather than later.
What are your thoughts on this reasoning?
Here's my example.
I'm running a Clickbank diabetes product in Mexico on MGID. I'm getting cheap clicks, so I figured this was good to test. I'm using a funnel with a catchy ad -> you must watch this video lander -> VSL offer page.
The funnel needs to be optimized. However, I'm getting plenty of clicks to the offer. According to Clickbank, 2477 clicks. I have one conversion for $19.36. 19.36/2477 = $0.0078, less than a penny per click to offer off of my lander.
Multiply this by my CTR (after I optimize) and I'll need to be getting extremely cheap traffic to make this campaign profitable.
Is this a good time to quit? 1. There aren't many Mexico diabetes products to test to see if there are ones that convert better. 2. I know that the earlier steps in the funnel affect offer performance, but is it realistic to expect these steps to bring the offer performance to a point where this campaign has the potential to become profitable?
It's always good to be mindful of the earnings per click to offer and cost per click to offer.
Some sites will cost too much per click to make the offer worthwhile, so you can cut early on rather than waste your budget spending 2-3x CPA on a site placement.
However it's still just one metric in the overall picture. It doesn't really become more valuable until you know how the offer and traffic source performs.
You didn't mention how much you spent on traffic? It's easy to get deep in the red starting out on native. You should look to start with a smaller set of data than trying to run the entire country of Mexico on MGID if that's what you're currently doing.
Thanks for your reply. I've spent about $80 and got one $20 conversion. I'd be optimistic if it weren't for the low EPC of the offer.
Originally Posted by thedudeabides
Originally Posted by p rand
That's not that bad. You can't expect the EPC of an offer you've just started testing on a new traffic source and targeting to be good - there are too many low quality sites dragging that number down, not to mention bids can be adjusted etc. A single conversion or two could change the entire outlook of a campaign, so it's too soon to call it quits.
How narrowly targeted your initial campaign is really depends on how much can be spent per day on your desired targeting and the number of placements. If there's say $1k/day mobile being spent for mobile in Mexico, but 1000 placements, then starting out by running every single placement is not an efficient use of your spend while you're trying to optimize multiple things at once. You're going to end up losing a lot of money, and probably end up making poor decisions based on which way the wind blows in a given day for site placements since you'll only be getting a small piece of that pie at a time.
Before you scale you need to figure out the best performing ads, lander, & offer, geo if applicable as well. That's better done with a smaller set of traffic and placement list that's still representive of the whole. Ask your rep what the best performing sites are in Mexico and start with those perhaps.
Also I'd definitely test desktop too with VSLs.
Home > Paid Traffic Sources > Native