How To Do Conversion Rate Optimization On Your Amazon Niche Site

By Last updated on:
CRO for affiliate sites

Let’s say that 100 people visit your site every day.

You have a 5% click-through rate.

This means 5 people click a link and land in Amazon’s magical world with your affiliate cookie stamped on their feet.

But here’s the problem:

What about the other 95 people who get to your page and don’t click?

Is it you?

Is it them?

Has Thanos snapped his fingers again?

What drives them away?

How can you get them to stay?

And more importantly, what can you do so they click your link?

Figuring out what happened to these 95 people and getting them to engage with your page is what Conversion Rate Optimization (CRO) is.

And if you care about increasing your revenue, learn how to do it now (or get someone to do it for you).

So why should you even care about CRO?

Because it’s crazy not to.

You make more money.

Your site visitors have a better experience when they land on your page.

And your business grows.

Everyone’s happy.

You already spend money and time to get people to your site. Don’t you want more of them to click your links?

With CRO, you systematically find out what you need to do to get more people to click your affiliate links.

No guessing or black magic. But pure scientific data that you can replicate and scale.

And the more of them that click, the more money you make. The same traffic. but more moolah in your bank account.

Now isn’t that a clever way to run your Amazon niche site?

Sounds good?

So let’s talk about the preparations you need to get ready for battle!

What to work on: The 80-20

First, you need a website with traffic that already converts

Then do this:

1. Find the pages that give you the most revenue and have the most traffic

    • I’ve found that for many affiliate sites, 20% of the pages usually give 80% of the revenue. Work on these pages first. With more traffic, you’ll see results quicker.

2. Work on above the fold content

    • A good percentage of visitors to money pages are ready to buy.
    • But what happens as soon as they land on the page?
    • A big long block of text.
    • With a drab, ordinary and forgettable call to action underneath it.
    • This is like having a store and hiding the cashier in the stock room!
    • Don’t do that. Work on optimizing above the fold content first.

How to decide which element to test

At some point when a visitor lands on your page, he gets stuck.

Maybe he has a question that your page doesn’t answer…

Maybe he can’t find a convincing reason to click the link…

Maybe he’s a bit confused and can’t be bothered to figure things out.

And when he’s stuck,

He leaves.

So your job is to find out what it is on the page that gets your visitor in this muddle.

And then unstick him.

This usually means optimizing your design and/or your copy.

Mobile may be your Achilles heel

people on cell phone

Have you recently looked at your site’s stats to see how much traffic comes from mobile?

I have, check it out!

Mobile site data

55.9%!

And this isn’t an isolated case.

Many of my other sites have stats that look like this.

But what I’ve noticed on many of the sites we’ve worked on is that the design doesn´t have the mobile user in mind.

When I tell people this, they usually respond by saying that their website is responsive.

But that’s not what I’m talking about here.

Being responsive and a good converting page are two different things. You should be looking at your page and asking,

“Is the mobile design doing exactly the same thing it’s supposed to do on desktop?”

If the answer is no, then it’s time to make some changes like these:

  • Make a big call to action button that’s easier for people to click.
  • If you have a comparison table, have all the data in one column, so users don’t have to scroll sideways.

Pro Tip:

When you do your CRO test, separate mobile, and desktop conversions. These could be starkly different from each other.

Your comparison table is failing you

Often, especially on money pages, you can increase conversions by getting out of the way of the people who are ready to buy.

They don’t want to read a long wall of text. They’ve done most of the research already. They just need a quick scan to:

  • find the best product that fits their requirements
  • see where to buy the product

The best way to do this is by using a comparison table.

In many of the tests we’ve done, using comparison tables has a positive effect on conversion rates. So I think it’s one of the first tests you should do to increase conversions.

I’ll tell you what though:

Many affiliate websites use a table plugin.

From what I’ve seen, these tables are clunky and not user-friendly.

So my suggestion is to use custom code. If you don’t want to learn how to do it, then at least hire a coder to do it for you. It’s worth the expense.

sample html code
Now once your table works on both mobile and desktop, give it the power it deserves by making the following changes.

  • Make it easy for visitors to compare products. If they have to think too much to compare them, they get stuck. And you know what stuck people do online. They don’t get glued to the page. They click the back button.
  • Target different demographics who are interested in what you’re selling. What do I mean? Different types of buyers land on your page. One wants the cheapest product, another wants something the whole family can use, and a 3rd one wants all the bells and whistles. Think about this when you list the benefits and address their different buying motivations.
  • Read Amazon reviews and take the language that people use to describe the product. I’m not saying copy the review word for word as that’s against Amazon’s TOS, but pick out a few words that people regularly use in their reviews.
  • Don’t put too many things in the description. 3 is ideal, but you can put as many as 5
  • Test which converts better: no pricing or pricing (make sure it’s attached to Amazon API to follow TOS)
  • Use clear, crisp images. In many of our tests, using an image almost always increases site conversion.
  • Test having less text on the mobile version of the page. You can hide some elements by using CSS.
  • Put your table where it’s easy to see. I’ve seen a lot of great comparison tables hidden way down the page.  Don’t make it difficult for visitors to find your Amazon affiliate links.

Your call to action isn’t doing its thang

This is the easiest one to change.

It’s low hanging fruit.

This usually includes changing the:

  • text
  • button size
  • button color

Here are some phrases that you can test:

  • Read Amazon reviews
  • Buy now at Amazon
  • Check price
  • Purchase now
  • Click here for latest pricing
  • Buy Now

As for the button,

I always think of it like designing for a 5-year old. Make it prominent and bright so that it’s easy to see.

Some things to test:

  • Make the color different from all the other colors on the page so that it directs the eyes toward it.
  • Test whether bigger or smaller buttons work

Your sidebar is getting too much attention

People love sidebars.

But a lot of times, especially on a page where your goal is to direct people to Amazon, sidebars are a distraction.

It adds an additional choice for the visitor.

And when there are too many choices, the overwhelmed visitor leaves.

So test not using it. You might be surprised by the results.

Your page does not feed the scanners

When a visitor first lands on a page, he scans.

He quickly checks if he’s in the right place.

This happens in milliseconds, but it influences a person’s emotional and psychological attachment to the page and your site.

So what can you do to make your site better for scanners?

  • Don’t be scared of white space.
  • Add lots of images and videos.
  • Use bullet points.
  • Have a clear headline that shows what the page is about

Pro Tip:

We’ve seen in many tests that a lot of people click on images. So why not attach your affiliate link to the image?

The CRO template that drives conversions through the roof

Gotcha!

Sorry to burst your bubble. But there’s no such thing.

See, I found out early on that, generally, people like quick fixes.

They need a template of the elements that have resulted in high conversions for us so they can use it on their sites.

So to satisfy people’s curiosity, we did a test.

We tested 10 different pages in different niches and used the same template.

The results?

Conversions increased on 3 pages, decreased on another 3 and 4 didn’t make any significant change.

See?

There are no shortcuts to CRO.

People react very differently depending on the niche, the season, their personal situation, even the day of the week.

That’s why we always test.

A/B Tests

An A/B test gets rid of guessing and makes everything data-based.

This is the heart and soul of conversion rate optimization. It approaches the process as a scientific data-driven exercise.

In an A/B test, you have two different versions of a page. You show one version to half of the visitors and the other version to the other half. One is your control (your old page), and the other one is your variation (the page with a slight change).

You pit these pages against each other and see which one converts better.

You’ll need tools

If you find the bit on scientific data-driven tests intimidating, don’t be.

There are tools that do the calculations for you and make your life easier. These are

  1. Heat mapping software
  2. A/B test software

Heat mapping software

Wouldn’t it be great (& creepy!) if you can stand behind every visitor to your site and record exactly what they do?

  • Which link do they click the most?
  • Where do they hover a lot?
  • Until what part of the page do they scroll to?
  • How long do they look at the images?

Guess what? 

You can.

Not literally.

But with a heat mapping software.

Install it on your site, and it will track user behavior.

This is what it looks like.

The red/orange color shows the part of the page with the most engagement and activity.

The blue, less so.

How does this help you?

Well, it helps you decide which element to work on first.

For example,

  • If they hover a lot on one link, you can test that link first. Maybe change the text. Maybe turn it into a button.
  • If most visitors drop off after a certain point on the page, then don’t put any of the important elements like the comparison table below that.
  • If there is an image or text that visitors click on but is not linked to Amazon, simply add your affiliate link. A very easy win.

Use one of the heat mapping software programs below:

  1. Hotjar
  2. Heatmap
  3. VWO

A/B testing tools

The most important tool you need is an A/B testing software.

This powers every CRO experiment.

It does all the hard lifting for you. It crunches up the data, records it, calculates the results, and tells you when to stop.

It’s your very own statistician and research whiz.

There are a lot of A/B testing tools to choose from. We use VWO in-house. But it only makes sense on enterprises that are running a bunch of campaigns.

However, for your first campaign, you can get a 30-day free account from VWO here

Let me walk you through the steps of conversion rate optimization

First, you need traffic.

This is traffic that has already proven itself to convert. Don’t just buy it to start your test.

Clear?

You also need enough traffic to make your results statistically significant. You don’t want it too small as the results will not be accurate.

Ready?

Buckle your seat belt. It’s going to be a fun ride.

1. Choose your page

Remember the 80/20 rule.

Find the page that makes you the most money and has a lot of traffic.

With an Amazon niche site, this would be a review page, a comparison article, or a list of product recommendations.

2. Install a heat mapping software program

Now that you’ve got the page install software for heat mapping.

Choose from the list above.

It doesn’t matter which one. Just choose one.

Here’s my tutorial on how to use heatmap.com if you need some help using it.

3. Decide which element to test

Based on the results of the heat mapping test, find the element on your page that gets the most activity.

Is it a link?

A call to action button?

The comparison table?

An image?

Choose only 1 thing to test.

More is not better.

You might think that changing many elements at once will save you time and give you better conversions.

But you could not be more wrong.

The main problem with this is that you won’t know exactly which element has changed the conversion rate. And when you don’t know this, it’s difficult to recreate the test on another page on your site.

So, just choose one.

4. Decide what to change the element to

Say you’re going to test a text link.

Think about the visitors who are not clicking that link. The ones who are not engaging with the page.

Now study the page and critic it.

  • Is it clear that it’s a link?
  • Does the link increase anxiety?
  • Do you think your visitors would prefer wording that tells them where they’re going? Like Check price at Amazon or would they prefer Check Price.

5. Know your baseline

You need a baseline to compare any changes to – a benchmark that you can use your new test on.

Hopefully, you’ve already got this data in your Google analytics.

If not, you will have to do the extra step of setting a click-through goal on the page. Here’s Matthew Woodward’s tutorial on how to do that

6. Form your hypothesis

This is a scientific process, lads.

So you need a hypothesis.

Glad you listened to your Science teacher, huh?

With a hypothesis, you know exactly what you’re testing, why you’re testing it, and what you hope to achieve.

Form your hypothesis this way:

“If X, then Y, because Z (your rationale).

So you can say:

 “If  I change the text on the button from Buy Now to Check Price at   Amazon, then the conversion will increase because it stops the visitor   from guessing where the link goes which lessens anxiety.”

7. Write your goal

Goals may differ depending on the page you’re working on.

To find your goal, decide what your experiment is trying to influence.

Why are you doing this test?

What do you hope will happen?

For an Amazon niche site, there are usually 2 goals. These are:

  • to increase affiliate income/revenue
  • to increase the click-through rate from the money page to Amazon

8. Start A/B test

It’s time to do your A/B test. You’re going to need a software for this.

The whole idea might sound daunting. But it really isn’t.

With the right software, you’ll be amazed by how easy it is. Just input the things you’ve done, and it will do everything for you.

You don’t even have to hurt any lab rats in the process.

If you’d like a quick tutorial on how you can use VWO for your first test,  here’s a video of me doing a button test on VWO. 

Once you’ve done this, it becomes a waiting game.

How long you wait depends on the traffic. For big sites with lots of traffic, it could be as short as 2 days. But we’ve had tests that lasted up to 6 weeks.

9. Stop the test

So when do you stop the test?

The quick answer is that your software will tell you.

So what exactly is your software waiting for?

Let me introduce you to the almighty statistical significance.

Say you go to a school with 500 kids.

You ask 5 kids which teacher they prefer: Mrs. X or Mr. Y.

4 of the 5 kids say, Mrs. X.

Can you conclude from asking these 5 kids that most kids in the school prefer Mrs. X?

No, you can’t. Because you haven’t asked enough people to get a result that is statistically true.

This is statistical significance, and when doing a CRO A/B test, you wait to reach it before you stop the test.

But as I’ve said, you don’t have to concern yourself with this as your A/B testing software will do it for you.

But if you really must, you can check out Optimizely’s Sample Size calculator to know how much traffic you need.

What Next?

Has your test finished?

Has it reached statistical significance?

The results will lead the way to your next steps.

Successful?

If your new variant wins, great!

All you need to do next is to roll it out to 100% of your visitors.

Then you have two choices to make:

1. Try the Same experiment on another high-traffic page on your site.

Notice I didn’t say to roll out the changes to all pages on your site.

No!

Each page needs its own test.

You see, it’s possible that what affects the visitors of one page may not be the same things that matter to the visitors of another.

2. Refine the Test

CRO is a continuous process.

Once you have a winning result, you can continue to refine it for higher conversions.

So do a new test on the same page and increase conversions even more.

This is the beauty of CRO. As you continue to refine your tests, your conversions will start to skyrocket because of the constant improvement over time.

But what if the test failed?

Even a failed test can give you insights into the psychology of your visitors.

Maybe the changes you made are of no importance to them.

Or maybe it increased on desktop but not on mobile (definitely worth checking!).

Don’t be disheartened when a test fails. It’s disappointing, but your experiment is not in vain.

Go back to the drawing board and do another test.

That’s right, soldier!

“There are two ways to do something: the right way, and again.”

Setting Expectations

Many CRO articles tell you about all the big wins.

You don’t often see a lot of stories about the failed experiments.

But the truth is, there will be many.

Lots of them.

And once you find a winning variant, you’ll win big time. You could end up making 100% more than what you’re making now with the same traffic you have. And all the failed tests will have been worth it.

So if you want to increase your conversion rates, then stick with it. This is the long game. And once you see a winning test, you’ll be glad you stuck it out despite all the ones that failed.

Conclusion

So there you have it.

The beginner’s course to conversion rate optimization for Amazon niche sites.

As you do your test, develop a system that you can use over and over again. You’ll be repeating the same things in many experiments on many pages.

When you do, you can easily go through all the same steps next time you do a test, and you don’t have to start from zero every time.

kurt

About Kurt

Hi, I’m Kurt. Founder & CEO of Convertica. I live and breathe conversion rate optimization. I hope you enjoy my findings. Interested in CRO? Let’s connect.

7 Comments

  • Reply
    Jeremiah Say
    June 11, 2019 at 10:26 am

    I am blown away by the amount of information in this post.

    I like that you recommended an A/B testing software.

    Since the beginning of the post, I was waiting for you to recommend a way to track clicks to Amazon (from which posts) because Amazon’s dashboard just doesn’t show where the links come from.

    Amazon basically only shows: “Okay, you have 200 clicks today, with a conversion of 2% and you are making this much.” That’s it!

    I don’t usually make full use of the tracking-code provided by them because they only limit to about 100 or so.

    So my question is, is there a way to track clicks (I want to know where the clicks are coming from exactly) without using Amazon’s tracking-code.

    I would love to hear your thoughts on this.

    Thanks,
    Jeremiah

    • Reply
      Andrej Ilisin
      June 11, 2019 at 11:06 am

      Hi Jeremiah, thanks for stopping by.

      I usually create a unique ID for my top 10 posts (traffic) and add it to the specific page so I can track where’s the traffic/sales coming from. You can go one step further and track page, table vs contextual, etc. Hope that helps.

  • Reply
    Jeremiah Say
    June 11, 2019 at 11:51 am

    Alright.

    I guess that’s what I’ll do too! Thanks for taking the time to reply.

    I appreciate it.

  • Reply
    delwarjahan
    June 12, 2019 at 7:41 pm

    Jetpack gives you a pretty good data on what pages are hot and how many times people visited those pages and exit links. Also, search terms and referrers.

    It’s not quite easy to determine everything from that but it helps.

  • Reply
    Steven
    June 14, 2019 at 1:01 am

    Great article with lots of useful information and advice! Thanks!

  • Reply
    Tom
    June 16, 2019 at 5:09 pm

    Thanks so much for the interesting post and the lonk to Matthew’s tutorial on tracking affiliate clicks (and other stuff) with Analytics. Helps me a lot.
    All the best,
    Tom

  • Reply
    Marty McLeod
    July 7, 2019 at 3:33 pm

    Nice article! Kurt really knows the game well. 🙂

    In fact I’ve done some A/B testing after taking his CRO Academy course. It’s really interesting to see how your assumptions regarding CRO ideas will turn out in reality – often they’re not at all what you expect.

    One thing I’ve seen over and over is that there are FAR too many myths about what “converts” well for affiliate sites. In fact a great deal of it is completely wrong, and I constantly see people losing out on potential earnings because of following bad advice.

    If there’s no data behind someone’s claim (or if they’re not really experienced with CRO, which is often the case) then there’s a good chance it’s poor advice and could even cost you earnings.

    One more important thing to note is that the click through rate (CTR) is not the same as conversions. In fact, I’ve had test results where clicks were the same or higher than the original post but resulted in lower conversions.

    People’s buying behavior isn’t always a direct correlation in the number of clicks. Kurt tells that for real CRO testing we evaluate the actual sales conversions & earnings, rather than just relying on clicks.

    That’s very important!

    Thanks for a great article. 🙂

Leave a Reply

Zero To 100K With SEO

A free 5-part SEO email course that will show you step-by-step how to grow traffic from zero to 100,000 visitors per month (with search).

Sign Up

Enter your email below to get the Zero To 100k course