This guide is for sites that are already getting good traffic and ranking for affiliate keywords.
The first thing you need to do is to know your numbers. This is very important. You can't test without a baseline. This is why amazon allows you to have over 100 different tracking IDs. There are two big metrics that I use: CTR (click through rate) and EPMV ( Earnings per 1000 visitors).
Usually I just test for CTR because that's the best thing to do for Amazon Associates or if you know which affiliate program is best. I test for EPMV when I am testing out different affiliate programs.
Testing for CTR
So let's start with CTR.
First, set up Google Analytics and let it gather data. Also make custom affiliate tracking IDs for your most popular pages. It's easy for amazon but for your private affiliate programs you will probably need to create a sub-ID. It should be easy, depending on the program.
Think of every popular page on your website as its own website.
After a few weeks you should be able to get a baseline number for CTR.
To get the baseline CTR:
Go into Google Analytics
- Set the time-frame from 2 weeks back, or less if the page gets a lot of traffic
- Click on Behavior
- Click on Site Content
- Click on All Pages
- Find the page you want and look over to the column that says 'Unique Pageviews'
- Copy that number and paste it onto a spreadsheet
Go into Amazon Associates or whatever affiliate program you use
- Set the same time frame as in Google Analytics
- Get the number of clicks for that specific page
- Copy that number and paste it onto a spreadsheet
Now divide the number of clicks from the affiliate program by the number of unique pageviews. That's your CTR.
Generally, if it's an amazon review page like 'best chainsaw' then you want at least a 30% CTR. 35% is really good. The good news is that if you are around 15% then you know that you can make improvements to probably double that number.
So get the baselines for all of your popular pages.
Now onto the actual testing part:
You are going to use the scientific method here. Think about what you think would improve your CTR on your page. For example, let's say we want to add a comparison table.
There are two different ways that I have carried out split tests, sequential split tests and the more common A/B split test.
Sequential split tests are when you gather data from one period, the control. And then directly afterwards, you test out the variation for the same amount of time then compare the results. For example,
- Two weeks with your control page
- Two weeks with the page with the comparison table.
- End the test and compare the CTR for both the control and variation.
If the variation with the comparison table is better by 10% then you keep it. If there is a big difference in CTR then it's safe to say that you can add a comparison table to other pages on your site that are similar.
This method of testing I've found to be good if you are making big changes, like adding a comparison table. It's not reliable for small changes like changing the color of the buy now button. You also can't do sequential split testing during the holiday season. Another bonus of sequential split testing is that it's free.
The other type of testing is A/B testing. This is when the the control and variation pages are shown randomly to visitors during the same time period. This makes it a more accurate way of testing. However you need to use either a wordpress plugin or a program such as Visual Website Optimizer (VWO).
VWO is $60 a month for the most basic plan. It's simple to use but it will take about a week to get used to how it works.
When you start doing tests make sure to keep a spreadsheet. If you use VWO then the program will tell you when the test is over (when statistical significance has been achieved). For a sequential split test it will require a more educated guess on your part. But usually at least a week if you have a lot of traffic.
Layouts That Work for Me
Here are product review layouts that I've found work best for me. Keep in mind that they work for my main affiliate site, but they might not work well for you because of different themes and different niches.
- Comparison Tables with Images
- Buy Now Buttons
- Text link in product descriptions
- Text link in first paragraph
- Hyperlinked Product Images
I usually avoid the text link in the first paragraph because I don't want to have too many affiliate links. Google can penalize sites with too many. I also make sure to 'nofollow' every affiliate link.
That's pretty much it. Let me know if you have any questions, and send me a PM if you want me to take a quick look at your site.
The first thing you need to do is to know your numbers. This is very important. You can't test without a baseline. This is why amazon allows you to have over 100 different tracking IDs. There are two big metrics that I use: CTR (click through rate) and EPMV ( Earnings per 1000 visitors).
Usually I just test for CTR because that's the best thing to do for Amazon Associates or if you know which affiliate program is best. I test for EPMV when I am testing out different affiliate programs.
Testing for CTR
So let's start with CTR.
First, set up Google Analytics and let it gather data. Also make custom affiliate tracking IDs for your most popular pages. It's easy for amazon but for your private affiliate programs you will probably need to create a sub-ID. It should be easy, depending on the program.
Think of every popular page on your website as its own website.
After a few weeks you should be able to get a baseline number for CTR.
To get the baseline CTR:
Go into Google Analytics
- Set the time-frame from 2 weeks back, or less if the page gets a lot of traffic
- Click on Behavior
- Click on Site Content
- Click on All Pages
- Find the page you want and look over to the column that says 'Unique Pageviews'
- Copy that number and paste it onto a spreadsheet
Go into Amazon Associates or whatever affiliate program you use
- Set the same time frame as in Google Analytics
- Get the number of clicks for that specific page
- Copy that number and paste it onto a spreadsheet
Now divide the number of clicks from the affiliate program by the number of unique pageviews. That's your CTR.
Generally, if it's an amazon review page like 'best chainsaw' then you want at least a 30% CTR. 35% is really good. The good news is that if you are around 15% then you know that you can make improvements to probably double that number.
So get the baselines for all of your popular pages.
Now onto the actual testing part:
You are going to use the scientific method here. Think about what you think would improve your CTR on your page. For example, let's say we want to add a comparison table.
There are two different ways that I have carried out split tests, sequential split tests and the more common A/B split test.
Sequential split tests are when you gather data from one period, the control. And then directly afterwards, you test out the variation for the same amount of time then compare the results. For example,
- Two weeks with your control page
- Two weeks with the page with the comparison table.
- End the test and compare the CTR for both the control and variation.
If the variation with the comparison table is better by 10% then you keep it. If there is a big difference in CTR then it's safe to say that you can add a comparison table to other pages on your site that are similar.
This method of testing I've found to be good if you are making big changes, like adding a comparison table. It's not reliable for small changes like changing the color of the buy now button. You also can't do sequential split testing during the holiday season. Another bonus of sequential split testing is that it's free.
The other type of testing is A/B testing. This is when the the control and variation pages are shown randomly to visitors during the same time period. This makes it a more accurate way of testing. However you need to use either a wordpress plugin or a program such as Visual Website Optimizer (VWO).
VWO is $60 a month for the most basic plan. It's simple to use but it will take about a week to get used to how it works.
When you start doing tests make sure to keep a spreadsheet. If you use VWO then the program will tell you when the test is over (when statistical significance has been achieved). For a sequential split test it will require a more educated guess on your part. But usually at least a week if you have a lot of traffic.
Layouts That Work for Me
Here are product review layouts that I've found work best for me. Keep in mind that they work for my main affiliate site, but they might not work well for you because of different themes and different niches.
- Comparison Tables with Images
- Buy Now Buttons
- Text link in product descriptions
- Text link in first paragraph
- Hyperlinked Product Images
I usually avoid the text link in the first paragraph because I don't want to have too many affiliate links. Google can penalize sites with too many. I also make sure to 'nofollow' every affiliate link.
That's pretty much it. Let me know if you have any questions, and send me a PM if you want me to take a quick look at your site.