How to test native traffic sources and compile blacklists
Affiliates are often left wondering as to how many clicks their campaigns are supposed to generate without bringing any leads before you can blacklist a certain platform. Today, we’re breaking down one of the methods of testing native ad campaigns and creating blacklists.
Where to start from
- The minimum conversion rate you need to secure in order to exceed the break-even point and not to plunge into the red. That is, we want the traffic we are running to demonstrate a conversion rate higher than a certain threshold.
- How accurate should your blacklist be and what part of your budget you’re ready to spend.
The minimum conversion rate
The conversion rate (CR) is the number of registration forms (leads) divided by the total number of visitors. A conversion can actually refer to any desired action you want the user to take.
Not all submitted registration forms are approved and paid for by an affiliate network, but if you run sufficient traffic volumes, you will secure a stable approval rate.
On average, you get X for a single lead, where X refers to a commission you receive for an approved lead.
For instance, an affiliate’s commission is $500 ( all the numbers are given for the sake of convenience), while the approval rate accounts for 70%. This means that we will receive an average of $350 ($500 * 0.70) for a single lead.
CPC (EPC, RPC, etc) is the average price for a visitor (click, user). It can be defined either as the average income from one click (EPC — earnings per click) or as the cost per click.
Income from a visit: EPC = an affiliate’s commission * approval rate * conversion rate. For instance, if your campaign generates a stable conversion rate of 1%, your commission constitutes $500 and the approval rate amounts to 70%, your EPC will account for $3.5 ($500*0.70*0.01).
Let’s imagine we have found the perfect balance: a CPC of $2.5 allows you to get sufficient traffic volumes and you want your EPC to account for more than $3.
CR = EPC / (commission * approval rate).
In this case, the minimum conversion rate you need to secure is 0.86% or 0.0086 ($3 / ($500 * 0.70)).
Accuracy and probability
What is the probability that one click will bring you a lead? Given a stable conversion rate of 0.86%, the probability will account for 0.86%.
What is the probability that three clicks will bring you a lead? One hundred clicks?
A little bit of theory: the probability that at least one of the independent events will occur is equal to 1 minus the probability that none will occur.
The probability that one click will bring you a lead is equal to 0.86%.
The probability that there won’t be any lead is equal to 99.14%.
The probability that three clicks won’t bring you a lead: 99.14% * 99.14% * 99.14% = 97.44%. The probability that one hundred clicks won’t bring you a lead: (99.14%)¹⁰⁰ = 42.16%.
So, given a conversion rate of 0.86%, three clicks will bring you a lead with a probability of 2,56%, while 100 clicks — with a probability of 57.84%.
Counting the clicks
The problem is to find out what number of clicks will bring you a lead with a certain probability.
The probability that n clicks will bring you at least a lead:
Suppose we need to test out traffic that is supposed to generate a lead with a probability of 90%. If it does generate a lead, then you can continue running it. If not, add this platform to your blacklist.
Based on the formula given above, we can come up with another one:
And the final formula will look like this:
So, given a conversion rate of 0.86%, if you generate 267 clicks, the platform will bring at least one lead with a probability of 90%, according to the formula:
If the CPC equals $2.5, we will need to spend $667 to find out whether the platform will convert or not with a probability of 90%.
90% is probably too accurate and too costly. To calculate it with a probability of 80%, you will only need 186 clicks and $465.
It is recommended to use a probability of 80-90% in your calculations. 95% is very high accuracy, but in this case you will need to spend quite a lot of money on testing. If you don’t have huge traffic volumes, you can focus on a probability of 80% or even less.
The template for defining the number of clicks on Google Calculator (with a conversion rate of 1.2% and a probability of 80%).
Этот текст скопирован с сайта https://zorbasmedia.com