How to Split Test Amazon Affiliate Posts & Links to Increase Earnings

One of the best ways to squeeze out more money from your existing portfolio of web properties is to optimize your existing content to increase earnings.

You could be shocked at how much money you are actually leaving on the table once you optimize.

What do I mean by optimize?

a) increasing the number of clicks to your affiliate offers

b) improving conversion of affiliate offers

c) finding new affiliate offers

e) creating new affiliate content (new sites or  new targeted posts)

With my affiliate offers (non-Amazon), I’ve been having some wild success the past month because I’ve found some new affiliate offers that are doing very well on a couple of my sites. I’ll share the results in my upcoming (April 2017) income report. However, I want to stick to Amazon offers for this post and talk about how you can track & split test your Amazon affiliate content (the pages with amazon links) to increase your overall earnings.

Make sure you read my general post about how to dramatically improving Amazon earnings first, for some general yet effective tips that can boost your Amazon earnings. I’ve used every one of those strategies in that post to help me make six figures a year in JUST amazon earnings. So they work.

This post though, I want to cover a more advanced topic — split testing to improve Amazon content earnings.

I’m going to give you some of the exact strategies I’ve used to improve my amazon earnings on my existing content.

I’m not hiding these methods sand then trying to sell you some crappy ebook like most of the other marketers in this niche do. I’m giving you the full package here.

This can also be used for NON-amazon affiliate content (i.e. affiliate offers) as well or to improve user experience (bounce rate, time on page, etc).

The Why & How

You can improve your Amazon Affiliate earnings by finding the best-performing pages, optimizing those pages, testing the changes to see if they improve earnings and if they do, repeat the same thing for other posts.

Most people simply throw on a general amazon affiliate id and forget about it.

However, if you use a single amazon affiliate id for your entire site, you are possible leaving a lot of money on the table because you will be unable to directly track how well specific pages containing your links are performing.

And you can’t test improvements out for that page that may earn you more money.

The solution?

To use multiple amazon associate ID’s to track specific pages (and even links), then using the split testing plugin Simple Page Tester to see if changes to that page improve earnings.

Let’s cover how to do this.

Step 1: Creating Tracking IDs for Amazon Affiliated Posts

One of the more effective methods I have found that lets me sniff out my best performing posts/pages (i.e. a specific review, best list, general post, etc) is by tying a specific amazon associate id to ONLY that page.

While this can be tedious IF you have a lot of different pages, it’s worth the effort because you can suss out exactly what posts on your website are earning you the most money and then go back to optimize/split tests elements on that page to improve the click rates and conversion rates.

Here’s how you do this.

1. Create a unique Amazon Tracking ID for the post you want to track

2. Create New Amazon Affiliate Links Using That Tracking ID

I recommend creating a naming scheme so you can organize links by site, type, and post target. This makes finding and using those links for the right type of posts on the right site easier.

For example, if I was going to create an affiliate link for authorityincome.com, I might do something like:

authorityincome_POSTTYPE_POSTNAME

So all my amazon links for authority income would start with authorityincome_

While you don’t need to do this, it helps differentiate on the Amazon affiliate link drop down menu and the report section the different sites’ affiliate links. It makes it easier to grab the right link for the right site and to organize the reports when you want to see how each links does.

You can easily use Amazon’s Text Strip to do this:

Do not use this id as a general affiliate id on your site. The idea is that you can look at this tracking ID on the Amazon report dashboard to see exactly how many clicks links containing that ID get, how many sales you get, and the conversion rate.

3. Change All Affiliate Links on Post/Page to the New Link

Now that you have created a unique amazon ID, create a new amazon affiliate link WITH that tracking ID and use it for ALL the amazon links on a specific post.

This means you’ll need to go back to a specific post, the post you created that unique tracking ID for, and manually change all amazon links you have on that page to the new link.

Optional: Use Thirsty Affiliates to Manger Your Amazon Affiliate Links for More Control & Ease of Use

I recommend you use an affiliate link manager like Thirsty Affiliates with the Azon Addon.

I use this plugin on ALL my sites.

The reason why is that it allows you to add all your affiliate links to the plugin and manage them there. You can also use the Azon Addon to directly search for Amazon products from within WordPress (you won’t have to create the links to a product in Amazon).

There are some additional benefits of using the Thirsty Affiliates plugin.

  • You can get the Thirsty Affiliates Geo Location Addon so visitors from other countries will automatically be sent to the right country-specific amazon store (scooping you up a few extra sales you wouldn’t have had)
  • You can easily change the target amazon page to point to a new page/product by editing the added link from the Thirsty Affiliates link manager, rather than having to go back to the actual posts containing the link and editing each one. This can save you a shit load of time.
  • You can get the auto linker addon to automatically convert keyword phrases to affiliate links
  • Affiliate links can be set to be no follow and to open in a new window
  • Link is cloaked and appears more integrated into your site
  • Can track links clicks stats by clicks & category with the Statistics Addon

What Types of Posts/Pages/Site Content to Put Unique Amazon Tracking IDs on?

Examples of the type of post to put this link on:

  • A specific ‘Best List’ you want to track the overall performance of
  • A specific product review
  • A specific post with amazon links
  • A TYPE of affiliate content you want to track (which can include multiple posts, if you use the same tracking ID).
  • reviews
  • best lists
  • comparisons

I like to really drill down into as much detail as possible so I can fine-tune individual posts and even specific links on specific posts. But this does require you to make multiple tracking IDs on Amazon and it can get a bit confusing. If you just want to track by TYPE of affiliate content, you can do this too.

For example, I use a single tracking ID for all the content on a specific site for:

  • all review posts
  • all general posts
  • all Amazon Native Ads
  • Amazon’s Mobile Popup

In general, I like to track individual best lists — so I create a unique id for each best list I create on a site. This lets me see exactly which best posts are doing the most money. I can then go back and see if I can squeeze out, even more, money by optimizing and split testing that post with some changes.

For reviews, I tend to use a single review id for all my reviews. However, you could drill down even more by making reviews id’s by category. For example, if you had a tech review website, a unique affiliate link id for:

  • mouse_reviews
  • keyboard_reviews
  • phone_reviews

This would let you track the performance for each category and you could see what category getting the most clicks and earning the most, and focus your optimization efforts there.

Step 2: Track& Monitor the Affiliate Links

Once you’ve inserted your specific links into the post/s you want to track, you now need to track how those specific posts are doing with Amazon.

If you don’t know how well a post is doing, in terms of clicks and revenue, you don’t know what to optimize. So this is absolutely key.

To track and monitor your post is simple, assuming you’ve done the previous steps.

Just go to the Amazon report dashboard and look at the Ordered Items and group by tracking ID. You can see exactly how each tracking ID (which, if you set it up as I’ve told you to do) will represent how your individual post tied to that ID is performing.

You’ll want to track the posts long enough to get some real data. Doing a day by day comparison may give you inaccurate stats.

I recommend tracking your ID’s for a full month so you can gauge things accurately.

You don’t want to run a test for only couple days and end up picking out a few posts that had a few lucky sales or clicks which skew your results or you could end up wasting time optimizing the wrong post.

Given that this process can take you a month or two, you don’t want to waste time on the stuff that won’t earn you the most for your ‘efforts’.

Note that you can gain another level of tracking stats by using Thirsty Affiliates with the Thirsty Affiliates Statistics addon. This plugin lest you analize your affiliate link click stats directly from WordPress and sort them by categories (if you’ve added the link to a made-up category) or by views or clicks. I use this to help sort what ‘types’ of affiliate links are doing best so I can focus on broad category types to optmize.

Step 3: Pick Out the Best Earning Content to Optimize

Once you’ve run your basic tracking for a couple weeks or a month (I recommend a full month), you need to identify the winners (and the losers).

The idea here is to look at what’s earning you the most and then optimize that content to earn you even more money.

What about optimizing the worst performing content?

You can — it could be there’s a lot of untapped potential with your posts/content that are under performing.

But chances are you’ll earn more money by taking your best performing content and making it perform better rather than trying to improve your worst performing content.

You can always, as time permit, go back to test & improve your worst content ONCE you improve your best performing content first.

How many posts should you test and optimize? Well, ideally you would do so on EVERY affiliate post you have to maximize your overall earnings sitewide. But given these tests are time-consuming and there’s a lot of effort involved, I recommend working on your best content first then going from there.

I recommend starting offer with your best 2-5 earning pages on a site and perform split tests (next section) on those first.

By improving a couple of your best-earning posts, you may significantly improve your overall earnings. I’ve found that 20% of my content earnings 80% of what a site makes on amazon. So focusing on that 20% is a more productive use of your time.

Remember: if you split test and find your prototype test improvements earn you more money on a couple posts, you can go back and apply that change to ALL your posts/content.

Step 4: Create Another Set of Amazon tracking ID’s

You’ve already created a specific Amazon ID to test how a post/page on your site is doing on Amazon.

If you want to optimize that post/page via a split test, you’ll need to create ANOTHER  unique ID to represent the split test change you want to test against.

So for example, if I make a ‘Best Gaming Mouse’ list. I will in Step 1 have created a post-specific tracking ID for amazon tied to that page.

Call it

authorityincome_bestgamingmouse-555

This tracking id is used for all affiliate links on my ‘best gaming mouse’ post.

I now need a variation id to put on my variation post that I want to split test. So I’ll now create another amazon tracking id that will be used on the new variation post.

authorityincome_bestgamingmouse_splittest-555

I’ll use the variation link on the NEW post that I make my changes/optimizations too. This id will let me know on the Amazon dashboard exactly how the variation post I’m testing against the original post is doing in terms of clicks and orders on Amazon.

Optional Step: Split Test the Posts

Once you’ve created a unique tracking ID and created the specific links and inserted them into the posts you want to track, and have monitored how those posts are doing, you can now look at optimizing.

I’m assuming you will or have already created the Amazon tracking ID for the variation post you are testing against the original.

There are several things we can do to optimize our posts to earn more:

  • Optimize the post content to increase our conversions
  • Optimize existing link positions and the link anchor text
  • Add in more links to the posts
  • Rearrange the post layout/design to increase clicks

Now the trick is to know if our changes actually improve things.

This is where split testing comes into play.

The principal is simple: compare two different versions of the same post and track which one performs better (in terms of #clicks and total earnings).

To make this work, you are going to have to do two things:

  1. use a split testing plugin that shows two different versions of the same post/page
  2. create a unique amazon tracking id for the new version of the post/page to track the performance (in my example in Step 4 it was authorityincome_bestgamingmouse_splittest-555)

Now, this A/B split testing is time-consuming as you’ll need to create the ‘B’ version of the post. The ‘B’ version is the post with the improvements you’ve made to the:

  • post design
  • link positions
  • link anchor text changes
  • link target (i.e. different product page linked to on amazon)
  • content changes

This modified post/page will be tested against the original  ‘A’ version.

How the Split Testing Process Works

The process of split testing is quite simple.

We’ll use a plugin (Simple Page Tester) to show two variations of the page — the original and the variation.

The plugin will randomly show one of the versions to each visitor who visits the page (they see one version ONLY).

Each version of a page will have on of the unique Amazon tracking ID’s we created.

The original post will have our unique amazon tracking id for that page ( in my example, authorityincome_bestgamingmouse-555)

The variation post (the post with our changes improvements) will have the variation amazon tracking  id (in my example, authorityincome_bestgamingmouse_splittest-555)

When visitors to either page version click on one of the amazon links shown on the page, their activity performed for that specific post/page will be recorded by amazon.

This is how we know which page version is doing best with amazon as we can compare how each post/page is doing by the results of each tracking id.

Once our split test is done after a month, we look at the two different tracking ID’s on amazon and we can compare them against each other so see which did better. We then know which version of the page has done best with Amazon.

We then know which version of the page has done best with Amazon.

How to Split Test Posts with Amazon

I’ve explained how the split testing process works using two amazon tracking ids. Now let’s talk about how to implement this in WordPress and our post we want to optimize.

First, you need the Simple Page Tester WordPress plugin.

I recommend the one-time fee premium edition because it calculates which split test version is the winner, in terms of traffic or conversions.

As we’ll be using Amazon stats ourselves, you’ll have to do look at the data yourself. But the premium version offers more and it’s the version I use. You cheapskates can use the free version though for this if you need.

You cheapskates can use the free version though for this if you need.

  1. Use the Simple Page Tester to choose your master page (your ‘A’ to tests against) which will have y our tracking amazon id on it.
  2. Use the Simple Page Tester to create your variation (with the variation having the alternative amazon tracking id on it)

Typically, I duplicate the original page and make the modifications to it.

Simple Page Tester will let you do this automatically for you.

Now you need to make the changes to the variation post (i.e. the new page you are making the improvements to).

Once you create your variation page (either duplicated from the master page or a brand new one), you will set it as your variation page with Simple Page Tester and start the test with Simple Page Tester.

The plugin will take care of the rest and automatically show the variation and the master page to new visitors.

NOTE: cache plugins can mess this up so for the duration of the split test DISABLE your cache plugin completely or disable the cache plugin for that the master page and the variation page. For example, I use WP Rocket and this plugin lets you disable the cache plugin for specific posts and pages. I simply click the checkbox to disable on on each page.

What to Do on the Variation Post / Page to Improve Earnings

Possible Effective Changes to Make That May Increase Clicks and Conversions:

  • Make sure you have a CALL to Action right at the top of the post after the first couple paragraphs. If your post is a review, put a nice big link close to the top of the post. If it’s a best list, give a section to your overall best pick right near the top. I typically like to put this after paragraph 2. You get more clicks like this, especially with the impatient readers.
  • Make sure you have a  CALL to Action right at the bottom of the post in the final paragraph. You’d be surprised at how many more clicks you get by having this final call to action.
  • If you have a best list, make sure you have a call to action right at the top of each item and right at the bottom of each item. You’ll have to play around with how you format it, but putting in specific links at the top and bottom of EVERY item on a best list can increase clicks.
  • Try Different Call to Action text. You’ll find different CALL to Action texts will perform differently. As a micro split test, you could even use ID trackings on the main call to action’s to split test WHICH one delivers more clicks. Some CTA suggestions you can try:
    • ‘Buy on Amazon’
    • Buy [PRODUCT NAME] on Amazon
    • Read Reviews on Amazon
    • Check Out the Reviews on Amazon
    • Get it on Amazon
    • Trying a CTA ‘Buy it on Amazon’ graphical button vs text link CTA
  • Include images of the item. For best lists, make a nice 300×300 or 500×500 image for each item and make that image an affiliate link too.
  • improving the post content. Better written content that seems less salesy will convert better. You’ll also want to play with the length of the content (short vs long posts). I’m not going to give too many tips here, as there is a great deal of variation in what will work and what won’t’ (which is why you’ll need to test) but you’ll have to play around with how you ‘write’ your content so it converts. This can significantly affect the number of clicks and also the end conversions.
  • testing posts with theme sidebar vs post without theme sidebar

Step 5: Split Testing Individual Links

You can split test whole posts/pages as I’ve given above. This is great for large changes to the post content itself, the position of the links on a post, adding images to the post, testing whether you get more clicks with a theme sidebar on your post or without one, adding in MORE links to the post content, etc.

But for a finer level of testing, you can track individual links in a specific post/page and split test at the link level.

Now you’ll need to choose what you want to track. As you likely have quite a few affiliate links to amazon in a given review or best lists, this is only really effective for the main links.

Examples:

Main CTA link at the top and bottom of a review

CTA Links after each item in a best list

If I go to the level of tracking an individual link, I’ll often make a broad change to the entire post links to test this. For example, tracking the CALL to Action text for all items on a best list.

For example, finding out if

‘Buy It on Amazon’ performs better as the CTA vs ‘Read the Amazon Reviews’

To keep yourself sane, I recommend ONLY going to this level of split testing ON your best performing posts. So if you have 100 posts and you’ve put in tracking ids to tell which posts are earning the most, I would pick out the BEST performing post. I would then test POST-wide changes. Then after I improve my results making large changes, I would then test out the link CTA with another split test.

Step 6: Choose the Split Test Winner

At this point you should have:

  1. Created individual amazon tracking ids
  2. Inserted the id into a specific post you want to track amazon performance
  3. Monitored performance for a week, several weeks, or a month to get some earning stats
  4. Choosen the best performing posts to squeeze more earnings out of
  5. Created a variation post/page to split test against using the Simple Page Tester plugin
  6. Run the test for a few weeks or a month to get data

Now you need to choose the winner.

How to Choose the Winner

Simple: if you look at the earnings for the variation tracking id you created in Amazon and it’s earned more money than the original tracking id (which is tied to the original post), then your changes earned more money.

At this point, you make those changes the NEW post by declaring it the winner of the split test using the Simple Page Tester plugin. Simple Page Tester will put your ‘loser’ in the trash by default, but you can ‘save’ your original post if you wish, just in case you want to revert to it, by simply RENAMING the post URL to something else to keep it around and removing it from the trash.

You can also look at some other stats besides just the amazon earnings, by using Google Analytics and looking at the user analytics for that specific post/page.

Compare the original page (the ‘A’) against the variation page (the ‘B’) with the following stats for the same period of time:

  • Bounce Rate
  • Time on page
  • Exit Rate
  • Page Views

Improving the bounce rate, time on page, and #page views may positively affect your overall SEO since it’s a sign that users are coming to your post and sticking around longer and / or interacting with the page more.

Indirectly, post that improves say your time on page, bounce rate, and page views, may increase your overall amazon earnings for your site. Perhaps someone goes to another post from the source post or clicks on more links to Amazon.

At the end of the day, I choose the post-split test version that makes more money in Amazon as the ‘winner’ of the split test, but I do look at these non-amazon analytics as well to help gauge if my changes have made for a better user experience.

I’ve had situations where I’ve split tested two versions of a  post and both did about the same on Amazon in terms of clicks and revenue, but one version gave a longer time on page and lower bounce rate (which I choose as the winner).

So you can split test for OTHER variables and not just amazon earnings.

Step 7: Repeat the Same Thing for Other Posts

Once you’ve done Steps 1-5 for several of your best posts, you can work your way down the list of posts by the most earnings.

So start with your top earning post, then do it for your second best earning post, then your third best, then your fourth best, etc.

I typically will do Steps 1-5 for my top 2-3 earning pages for an Amazon site.

After I see my change/improvement has worked for 2-3 posts, I will then broadly apply the changes to ALL my posts/pages or use the new template/change on future posts/pages.

This is how you can really maximize what you learned from the split test of a couple posts to really improve your earnings for all your posts of that type.

It’s important to note that you only want to split test ONE version against the SOURCE version. This m eans you are only testing TWO different pages/posts at the same time.

Trying to split test 3 or more pages/posts against each other can skew the data (you need a LOT LOT more data to tell which version is the best).

If you want to test multiple changes, I recommend doing so in steps.

  1. split test A vs B
  2. Choose winner (say B).
  3. Split test B vs a new variation (call this C)
  4. Choose winner (say B wins)
  5. Split test B vs a new variation (call this D)
  6. Choose winner (say D).

You do NOT want to create 4 page versions right away (A, B,C,D) then split test them against each other in the same test. This will really mess up your testing data. So keep it simple.

The Final Word

I’ve given you a very, very detailed look at how to split test your affiliate posts to earn more with Amazon.

This is a time-consuming process, but it’s powerful and you can significantly earn more money by spending the time to do this.

In fact, I do this process for EVERY Amazon site I have.

It’s possible to improve earnings by 10, 20, 30, even 50 percent or more.

While I’ve covered how to split test post & links for Amazon Affiliates, you can do so for non-Amazon affiliate offers too. You won’t be using the Amazon report dashboard or amazon tracking IDs. Simply replace the Amazon tracking IDS with your other affiliate links, making sure to make a unique tracking link for each post you want to split test / optimize  AND a variation of that link to test changes against.

I hope you guys learned something here.

I have a few other tricks up my sleeve to improve Amazon earnings and general affiliate sale earnings that I may share in a future post, but we’ll see.

Go make money…and happy easter!

Ben K

Leave a Reply