You’re about to launch a new affiliate campaign.
There are a ton of decisions you need to make, with one of the most important ones being which offers to test.
In a perfect world, you’d test every single offer out there. But you can’t do that because you’re limited by your budget and time.
If you’re interested in the car insurance lead gen space, you could have the option of promoting twenty different offers. You can’t test all 20 offers. Not enough time, not enough cash.
You need to be strategic.
Being on a smaller budget means you’re only able to test around three offers at first (in order to get statistically significant data).
The question is…how do you pick which three offers to test in the beginning?
Let’s talk about what most people do when they first start out.
The most natural way to make this decision is to see which offer has the highest payout. If there are twenty offers, you simply determine which ones are going to pay the most for each lead.
Another way for people to determine which offers to test is to look at metrics from the affiliate networks. One metric that affiliates love to use to judge offers is the network-wide EPC (earnings per click).
This is the average EPC of every single affiliate sending a click to that offer. In theory, it should work like Yelp reviews. The higher the better right?
These two methods seem to make the most sense…but I want to warn you that these are massively deceiving metrics.
I’ve seen people severely handicap themselves because they have self-imposed rules like:
- “I won’t run any offer that pays less than $5.”
- “I’m not interested in testing that offer because the network EPC is too low.”
I’m going to dive into why those mental models are wrong so you can make better decisions in selecting offers.
Why the Offer Payout Doesn’t Matter
It’s natural to think that the higher the payout, the better.
I mean, if a client wants to pay you $100 an hour, and another one wants to pay you $150 an hour, you should go with the person who pays you more. (Assuming all things are equal. The 2nd client could be an asshole and you don’t want to work with him.)
This mindset shouldn’t apply to affiliate marketing offers.
Let’s assume that you have the choice of promoting two different offers.
- Offer A – Payout: $14 per lead
- Offer B – Payout: $8 per lead
Most people would probably run “Offer A” and even skip the split test (*gasp*). There’s a 55% payout difference, so why would you waste your time?
It’s because there are different variables that go into that final revenue number.
Here’s an example to illustrate.
Offer A has a much higher payout than Offer B, but Offer B actually performs better in an even split test.
Remember that the PAYOUT is only ONE PART of the formula.
You also have to factor in:
1. Shaving / Scrubbing:
If you look at Offer B, it registers 1,100 clicks. Offer A registers 1,000 clicks. This can happen even if you send the exact amount of traffic to each other.
Scrubbing or shaving means the advertiser or network is purposefully not crediting you with clicks or conversions that you earned.
It’s a dirty thing, this industry.
2. Offer Page Conversion:
You should know this as an affiliate marketer, but not all landing pages are created equal.
Maybe the higher payout offer has a shit landing page. The traffic you send to it converts at a much lower rate, even though the offer is higher.
Side Note: Never judge an offer page by how beautiful it is. Sometimes ugly converts better.
3. The Database:
Maybe Offer A has been around for ten plus years, and they’ve built up a huge database.
What happens if JohnSmith@yahoo.com signed up for Offer A ten years ago and then he signs up again through your affiliate link?
Chances are that the company’s not interested in that lead. They already have it.
On the other hand, Offer B is a brand new company. They haven’t seen JohnSmith@yahoo.com yet and will accept the lead.
The networks could be using tracking platforms (cake, hasoffers, custom, etc.), and that can affect performance.
And the website servers hosting the offer pages? Their physical location relative to your target customer can affect conversions.
I’ve seen all these things happen countless times, with the result being a lower payout offer outperforming a higher payout offer.
Another way to think about it is that offers are kind of like basketball players.
Being tall is an advantage, but being the tallest basketball player doesn’t automatically mean you’re the greatest. Michael Jordan was actually below the average height for an NBA player.
Having a higher payout on an offer is an advantage, but it doesn’t automatically make it the best offer.
All that matters is the revenue you generate.
Why the Network Offer EPC Doesn’t Matter
Back in the day I use to choose offers based on the network wide epc.
Once again, this is the average of every single click an affiliate has sent to a network.
You can find this statistic either in the affiliate network platform when you’re searching for offers, or sometimes networks will send a weekly excel sheet with this data via email.
Remember…this doesn’t work like Yelp or Amazon reviews!
Take a look at this example.
Offer A has a Network EPC of $0.13
Offer B has a Network EPC of $0.44
I’ve seen people automatically think Offer B is better because it has a much higher Network EPC.
Take a few minutes to look at this and tell me what you can conclude from these stats.
Offer A has a worse EPC ON PAPER.
But that’s because Offer A has one person promoting it who’s an outlier.
Affiliate A is fucking up the stats by direct linking a massive amount of clicks to the network. There’s nothing wrong with their method of promotion, but it massively skews the data.
If you look closely, Offer A is performing better overall than Offer B if you don’t include Affiliate A’s statistic.
If you were one of those affiliates that choose offers solely based on Network EPC, then you would’ve ignored the superior offer.
What SHOULD You Do Instead?
Hopefully, I’ve opened your eyes to some new ideas.
All of this leads us to the next question: how SHOULD you pick offers?
Sorry, but there’s not an exact formula to follow. The only sure way to know is to straight up test everything.
But what’s practical?
I think the most practical way is to see what everyone else is running. The theory is there are guys out there with much higher budgets and more experience than you, and they’ve already done the split tests.
You can use a combination of spy tools and talking with your affiliate managers to see what these offers are.
Instead of asking for the network wide epc of an offer, you could ask your affiliate marketing “Hey, what are some individual affiliates seeing on this offer?”
Does this mean they will perform the best for you? Nope.
But it’s a good starting point.
The bigger takeaway from this article should be that there’s what “should work” in theory, and what actually does work in practice.
Featured Image by ksena32