Case Study: How We Got a 33.9% Amazon Click Through Rate with Stone Cold Traffic

Stephanie K   73

Updated on December 23, 2019

Here at the Thrive Themes marketing office, sometimes we go a little overboard.

We wanted to test just how powerful quizzes could be for an online business.

Do do this, we created a one page affiliate site consisting of nothing but a quiz, then sent a bunch of new visitors to it to see what would happen.

And since we're always up for a challenge, we made a stretch goal to see if we could make the experiment profitable.

If we managed to succeed, we would essentially be creating an automated money-making machine… resulting in the subsequent plan to move to the Bahamas and sip on cocktails for the remainder of our comfortable lives.

Either way, we were set to learn something.


So what happened to our retirement plans? Am I writing this while sunbathing on the shores of Paradise Island?

Nope. Not quite. But we discovered some amazing building blocks that could work towards a very successful strategy.

The only way we could have learned this was by testing, by doing something to find out where to pivot.

Luckily you don't have to start from where we did, you can pick up from where we left off. We’re going to show you exactly what we did, exactly what we did wrong, and how you can make use of all the good bits to increase profits in your own business.

Our Retirement Blueprint

If you’ve read any of our latest blog posts, you probably understand the power of quizzes for lead generation. What you may not have realized is how effective quizzes can be for affiliate and product websites.

A well made quiz on this kind of site is like having a digital shop assistant.

By asking a series of fun but segmenting questions, you can get your visitors to tell you exactly what they’re looking for. This gives you a chance to present them with the product best suited to their needs. In other words, the product that's most likely to get them to click buy.

We wanted to test this out. So this is what we did:

Made a ‘What Valentine’s Day Gift Should You Get Your Man’ quiz.

Drove targeted Facebook traffic to it.

Recommended products with Amazon affiliate links, based on the quiz taker's result.

We ran the quiz in the two weeks leading up to Valentine’s Day.

Let's Geek Out on Some Numbers

First, I want to show you the numbers from our experiment, so you can gauge if there’s a way to turn this strategy into something that might work for your business.

Later, I’m going show you the exact steps we took to put the whole experiment together, so you can reproduce the parts that make sense for you.

For now, here’s a quick overview of the Valentine’s Day quiz numbers:

Note: The Amazon click through rate here is referring to the percentage of people who clicked through to Amazon from the quiz results page (where the products are displayed). The overall click through rate, meaning the percentage of visitors that clicked through from Facebook ending up on Amazon, was 33.9%.

A Success or a Failure?

Certain parts of this experiment resulted in some very impressive numbers.

Clicks and Conversions for the Win!

86% of people who clicked the Facebook Ads completed the quiz and saw the results page. On top of that, 39% of people who made it to the results page ended up clicking through to Amazon.

That's huge! Especially considering this traffic consisted of people who were just happily browsing Facebook, with no intention of purchasing anything.

Amazon Conversions for the Win Solid Average

An Amazon conversion rate refers to the percentage of people who end up purchasing something after clicking through to Amazon on your affiliate links. The Amazon conversion rate from the experiment was 7%, which is average to slightly high for most niches.

So why aren’t we all in the Bahamas?

We identified two main factors preventing this strategy from becoming profitable. Depending on your situation and business, you may already have the solutions in place.

Your Cut of Each Eventual Sale

Amazon commissions are generally quite low. This, coupled with promoting such low price items (between $5 and $50), lead to some pretty lack luster figures.

Given the cost of the traffic we generated, our commissions per sale would have to be $29.54 just to break even. This means the value of the average purchase would have to be $445.74.

This is no easy feat. Especially considering the conversion rate would most likely drop with higher ticket items.

Increasing your value per sale can be done by selling your own products or by using affiliate programs that give higher commissions than Amazon. When relying on paid traffic, these two options are probably the only ways to make the strategy profitable.

Getting Enough Eyeballs

Based on our experiment, the cost per visitor was the biggest factor in making profit unachievable. Each new visitor cost us $0.70 in Facebook ad fees.

If your site is already drawing a steady flow of new visitors, you essentially cut out all of the experiment's expenses. Having this kind of organic traffic makes it a lot easier to profit from your quiz.

But don’t go counting your chickens just yet. If you use the exact numbers from our experiment, the model still isn’t overly profitable.

Say you have 1000 new visitors to your site each week, and around 500 of them take the quiz. Running exactly the same quiz, 433 would end up on the results page, 146 would click through to Amazon, resulting in roughly $14.54 a week in Amazon affiliate commissions.

$14.54 a week is nothing to write home about. But at least it would be $14.54 profit.

The Winners in this Spaghetti Western

The sweet spot lies in combining the two; access to cheaper (or free) traffic, and higher commissions per sale.

If your business involves any of the following characteristics, chances are you can turn this experiment into a profitable strategy:

  • A niche involving higher-ticket items, combined with existing traffic that’s ready to buy (such as a product review site)
  • Using affiliate programs with higher commission rates than Amazon
  • Selling your own products
  • Access to a steady stream of new traffic from unpaid or existing sources​

Bonus Benefit

It’s not just about the sales either. If you make your quiz results sharable, the boost in social shares can be very beneficial for your site. Google sees social shares as a website validation factor. Increasing social shares can help increase your search engine rankings.

Not to mention your quiz being shared across social platforms creates a new organic traffic source. It's often worth running a ‘just for fun’ quiz on your affiliate site, purely for the social share benefits.

Making the Non-So Money-Making Machine (In Less Than a Day)

Perched at my usual spot in a local café, armed with this crazy idea and the entire team’s hopes of the Bahamas resting on my shoulders, I set out to make our automated money-making machine.

Here’s exactly what I did:

Making the Quiz

I used Thrive Quiz Builder to create the quiz. The reason Thrive Quiz Builder is such a game changer for this type of quiz is because you can show different products based on the quiz takers' results.

1. Identify a Topic Based on Your Quiz Goal

First of all, you need to establish what your quiz is going to be about. 

If your goal is to increase social shares, choose a fun topic interesting to your niche. For example, on a baseball bat comparison site, a good quiz topic for social shares would be ‘which famous baseball player are you,’ or ‘only true baseball fans can get 100% on this quiz, can you?’ Fun and infinitely shareable.

If you want to know more about creating an irresistible quiz check out this post about what makes quizzes go viral.

If your goal is to increase affiliate sales, a quiz topic like ‘the ultimate [product] selector’ or ‘get the perfect [product] for your needs' would be more appropriate.

2. Define the Results and Products

Choose the products or results you want to drive your quiz-taker to. For the Valentine’s Day post we identified six different man types. Here’s an example of the results page for the ‘Modern Man’:

Each button and image linked to Amazon through an affiliate link.

We used the Dynamic Content feature in Thrive Quiz Builder to show a different selection of gifts depending on the result.

For the ‘Sports Fan’ type, it was ticket stub scrapbooks and LED helmets. For the ‘Man-Child’ type, it was VR headsets and magnetic putty (which actually sold surprisingly well).

Of course, you don't have to create 'man types' or groups of products like we've done.

Your results could just be the products themselves. In the baseball bat example, your results could simply be the ‘Easton S50 Youth Bat,’ the ‘Louisville Slugger Prime Bat,’ and the 'DeMarini Voodoo Youth Bat.'

3. Build out the Quiz

From here you can create the questions. Our questions were very basic, but we made sure to use images in the answers to keep the quiz-takers engaged the whole way through. Our goal was to get them all the way to the results page.

Finding images for the answers is easy. Just head to a free stock image site such as Pexels or Pixabay.

I got all the images for this quiz from Pexels. The quickest workflow I found was to download the images at a custom size, 200 pixels wide.

This saves you from having to resize and compress each image. Pexels is pretty good at giving you the smallest file size possible.​

​The only thing left to do was to customize the Splash Page template (the introduction page to the quiz).

We kept ours simple, but I’d recommend using Thrive Quiz Builder's built in A/B testing feature to test a few different Splash Page variations.

You don’t need a Splash Page, but having one allows Thrive Quiz Builder to get the initial statistics on how many people started the quiz. This data will show up in your Flow Report to give you a better whole quiz overview.

Generally there’s a drop off at each quiz step, so you’ll need to weigh up if this data is worth it for you, or if you’d like to lead visitors straight into answering questions.

Driving Traffic

In order to run a real experiment we needed traffic. Since we were running the quiz on a previously unused site, all of the traffic had to be paid.

1. Create Ads and Images

We chose Facebook ads as our medium for paid traffic, and made up some quick images and ad copy to drive clicks.

We tested different image and text combinations with AdEspresso. Alternatively, you can use the native Ads Manager within Facebook, which is free (minus the cost of the ads).

These were the two winning variations:

Again, these images were just from free stock sites. The text was overlayed using PicMonkey, which is a free photo editing program.

2. Choose the Right Audience

This is one of the most important factors in getting your cost per click down.

We went pretty general with our audience targeting, but made sure to include:

  • Women in a relationship
  • Women living in the USA
  • Women who liked female focused magazines and quiz sites such as Cosmopolitan and Playbuzz.

This assured the women who saw the ads were familiar with these types of quizzes and (one would assume) enjoyed taking them.

3. Drive Paid Traffic

Over 11 days, from the 1st of February to the 12th of February (two days before Valentine’s Day) we spent a total of $709 on Facebook ads for the quiz. This lead to 32,224 impressions and 1,014 clicks.

Click to enlarge

Our overall cost per click was 0.70c, but on the winning variations it got down to 0.39c and 0.47c per click.

This is important to note because as you hone your ads and your targeting, your cost per click will get cheaper and cheaper.

These definitely aren’t the best results you can get on Facebook, but the execution was rough on purpose. We wanted to see if we could create this money-making machine quickly, and with the average person’s resources and level of Facebook marketing knowledge.

Get Yourself to the Bahamas

While the experiment didn’t turn a profit, through testing, we learned the exact metrics we’d need to improve to make the strategy a success.

Given enough time, tweaking and testing, eventually our dream of the Bahamas could have become a reality.

Instead, we’re going to get back to work and pass the baton on to you. Take what we’ve learned (and lost) and apply those lessons to your own business.

There’s no denying the quiz numbers are impressive. From the 1,014 clicks onto the site (comprised completely of cold traffic), 33.9% ended up clicking onto Amazon.

Now it’s just about finding the right way to utilize this powerful tool in your own business.

Do you have any ideas? Do you think this strategy could work for you?

Let us know by leaving a comment below!

by Stephanie K  April 21, 2017


Disclosure: Our content is reader-supported. This means if you click on some of our links, then we may earn a commission. We only recommend products that we believe will add value to our readers.

Enjoyed this article ?

You might also like:

Leave a Comment

  • This is an awesome model and experiment. Thanks for sharing and trying new things, I always get new ideas from you guys. But I think there’s a typo…you ran the test 1-12 February, right? The article says 1-12 January.

    • Thanks for the heads up Todd, I’ve amended the post. You can tell how exciting my love life is… I don’t even remember when Valentine’s Day was :p

  • Thanks for the info. If you decide to do a sale at $47 on the 5 site version again let me know, when it came out i was away and missed the option.

    • Hi Keith,

      It’s a shame you were away! Unfortunately it wont be at that price again. Our motto is never to sell the product cheaper than it’s been, to be fair on those who bought it early. Shane has a really great post explaining the reasoning behind this.

  • Dear Stephanie

    Thank you for a very worthwhile experiment. I’ve a tendency to lack the courage and the money for such playfulness. I could though see a way to make it work for a high margin offer.

    Who knows, perhaps this is enough to get me to have a go and take a risk.

    I tend to use Pixabay for images and will add Pexels to my list.

    Please keep these coming, I really enjoy them.

    • Hi Steve,

      The similar experiment could also be done with a less violent add spend, provided you had more time to run and tweak the FB ads to get that CPC down. We were in a bit of a rush to promote it before Valentine’s Day.

      We love Pixabay too 🙂 There are great Creative Commons stock image sites popping up all over the place now. Pexels is my favorite because of that resizing feature though.

      Happy experimenting!

  • Love little experiments like this … I’m just about to run one for my son … he’s a youtuber and wants to grow his channel and his list. So, I’m doing a quiz, but I’m combining it with an online competition … the more people share the quiz with their friends, the more chance they have of winning the competition …

    … and I’m just going to start with some adsense on his thankyou page, but will also start selling some relevant merch through a print on demand company, so I only pay once the customer pays.

    Anyway … that’s my experiment planned for my son’s birthday which is coming up in July … he wants to be an online entrepreneur like his Dad and so what better gift to give than an experiment like this!

    So, my idea is: combined quiz with contest based on sharing in order to maximize organic traffic

    • Hi Nic,

      I love this!! Thank you for sharing. What a great Dad 🙂

      Social Sharing is another one of those perfect uses for a quiz, especially when you combine it with something like a competition.

      If you remember, come back and let us know how it all went!

      • That is incredibly awesome Nic, I would love to pick your brain about this more if you would be up to it.

      • Stephanie K,

        Thanks for a great article. It really helped get me wondering how I could use a quiz profitably or just to build my community (list).

        I’m also curious about how Nic is going to track his social shares.

      • I’ll let you know … especially now that you also have tagging based on answers and so I can use email to follow up specifically.

      • Ah … the magic of the internet … I use a stack of various softwares that do this. But combining the quiz with a competition is all theory right now for me. Best thing to do is to implement the idea … see how it goes … and then share the results. You can bet that if the results are great, I’ll be using it as a case study to promote Thrive Themes to my list.

  • You guys are full of clever ideas for imaginative posts. I look forward to them and am rarely disappointed. Keep it up!

    • You should see the ‘blog ideas’ list, we’ve got plenty more where that came from 🙂 Glad you enjoyed it Richard.

    • Hi Erik, this was more of a dedicated test to see what we could achieve with quiz builder and affiliate links. It’s also best to stick to one main goal with these quizzes, so for this experiment it was getting clicks through to Amazon.

      I agree getting emails would help eventually turn a profit, but then you risk distracting the visitor from the primary quiz goal of clicking through to Amazon.

  • Thanks Stephanie. This was a really interesting study. I’m looking forward to testing the concept in my own business.

  • Great article and really fun experiment. As a long time Amazon affiliate, I can say that you suffered not only from low commission rates but the thing that most people don’t know about the Amazon affiliate program is that their tracking cookie only lasts for 24 hours. Most other affiliate programs are 30 or 60 days, but with Amazon if the person comes back and makes the purchase 2 days later, even if they were completely persuaded by your great marketing, you won’t get credit for the sale (unless they added the item to their cart within the 24 hours, then you have 90 days for that item only).

    • Hi Jon,

      Amazon is an interesting can of worms. I know there’s a strategy about using an ‘add to cart’ link rather then just a ‘re-direct to product’ link, but there’s still a bit of controversy surrounding that. This article might be of interest, it goes into it a little further.

  • Hi Stephanie
    your post couldn’t come at a better time since I’m planning exactly this. Now there is one thing that I’m a little unsure:

    I know that Amazon is extremely sensitive with their affiliate terms and basically they want all affiliate links to be visible so that they can ‘theoretically’ track them all. So no links in ebooks, emails, even popups, etc. are allowed.

    Now my question is, where did you put the affiliate links? At the end did you forward the reader to a landing page? Is the link to Amazon in the last quiz result page?
    Can Amazon ‘again theoretically’ find these affiliate links without doing the quiz?

    Again I love to do the same thing but am very careful with Amazon Affiliate links so any information is much appreciated.

    Thank you!

    • Hi Thomas,

      Good question. We put the affiliate links on the results page of the quiz. We used the proper disclosures, but as for your question, I would ask Amazon Associate Support to be 100% sure. I would just be assuming from what I’ve read in their policies. It would be better to get it straight from the horse’s mouth.

  • What are the known compability issues and bugs with Thrive Themes. It sounds interesting. However, so does Thrive Content Builder until the reality that we are now 2 weeks into debugging and still unable to get it to work or to receive support other than generic, disinterested and unhelpful comments.

  • As always youre awesome..
    Thanks for your ideas and your inspiration!
    Keep up the good work,we need you 🙂

    • Thanks David. It’s a lot to take in. I tried to add as many helpful diagrams as possible to break it all up :p

    • I’m not going to lie Martin, they’re a ton of fun to set up. That’s why I keep putting my hand up to do these case studies 😉

  • Did you create a brand new Amazon affiliate account for this experiment? I have no way of determining how many clicks we have through our landing page to Amazon because we are already running Amazon affiliate promotion throughout our site. Do you use link shortener to track?

    • We had a dedicated account for the test. I know what you mean though, we’re doing a case study with an existing affiliate site and having the same problem. Especially because link shorteners are apparently frowned upon. If you figure out a good solution, make sure you post it here 🙂 I know you can add tags to links within Amazon, which can be used for tracking in some circumstances.

  • Whenever you purchase traffic and have to off-set that with a sale you are in my opinion playing with fire. The first thing I did was look at your “spend” and compared it with your “return”. I believe you are right when you say you would look for an affiliate offer that provided a higher return, however, I think the obvious move here is to somehow gather visitors email addresses in this process. Then your “spend” can be called “customer acquisition cost”, and you now have the opportunity to go in more directions that are perhaps more profitable. As always, I do appreciate your constant testing to try new things and your efforts to make things better.

    • Hi Grant,

      I agree, in the ‘real world’ this would definitely be a strategy worth implementing. But like you said, unless you make one of the changes suggested in the article, I wouldn’t recommend running our exact strategy even with the email capture. It would still need either cheaper/free traffic or a higher payout to get those margins looking a bit better, even with the added benefit of customer acquisition (otherwise that’s a very expensive customer acquisition cost).

      • Stephanie K,

        Recognizing the importance of getting the people to Amazon but also the long term benefit of getting a subscriber, couldn’t you, only on the results page promise a free bonus list of suggestions to enhance their romantic evening that would go well with their gift, and then have a list of theme related ideas, like for the baseball enthusiast: an evening watching one of the popular baseball movies, have a four course meal with each course being a base, etc. (Not being a baseball fan, I think I just ran out of G-rated ideas.)

        As people generally are not willing to make the effort to be creative,I would think such a list would be well received. The other side of this point is that it’s only worthwhile for a site actually trying to develop future sales with that kind of list, not a purely testing site like you used.

      • Hi John,

        Completely agreed. It would depend on your business model, and if repeat sales were a concern. If not, it’s not worth distracting them from the click to Amazon, but if it is, your suggestion is a great strategy. Especially if after opting in they stay on the same page, giving them another change to click on the products.

  • Hey awesome case study.

    I have some Problem with the Quiz Builder. I’m Wondering if it’s possible to Redirect to a Custom page instead to the result page.

    Quiz ==> Optin Gate ==> Custom Redirect ( instead result page)

    I know we can customize the entire Result Page , but sometimes we Need as marketers to Redirect to third Party destination too.

    It would be awesome if you can intergradre this to Quiz Builder

    • Hi Alejandro,

      You can achieve that by using the Results Page as an Opt-In Gate. You add an opt-in form on the Result Page, with or without the dynamic content, which can be set to redirect the user to whatever URL you want. So it would go:

      Quiz questions > Results page with opt-in > Successful opt-in redirects to a URL

      If you want to sent them so different pages based on their results, you just need to use dynamic content for the opt-in forms.

      Hope that helps!

  • Nice to see experiments done out in the open 🙂
    Thank you Thrive Themes for posting reality and not just theory!

    There are a few ways this may have been profitable.

    1. CPA offer to those completing the quiz that DID not click to Amazon.
    So if you gave a leave option on the page for those not purchasing, that link could take them to a CPA.

    With 880 Completions and 344 clicks, that leaves you 536 people leaving the final page. Convert some of these and you could bring up your income and reduce your loss, maybe even turn a profit.

    Alternative would be to add the click away to a list by offering something of value to them for completing the quiz but choosing to leave. This way you do not disrupt the click through to Amazon, but potentially capture some subscribers on those who leave?

    2. If you are running something like this for clients in the real world, then this could turn into leads for them and profit for you depending on their payment per lead. If this is part of a service you offer marketing clients, and you turn a profit, then obviously they will retain your service.

    3. As you mentioned previously, selling your own product could swing this in a big way. CPC of $0.70 can still be profitable, especially if they are entering a funnel process.

    With 24 sales on Amazon around the $18.50 mark, a product you sell at $30 of your own would see a break even on the front end and is still a low entry level purchase price.

    Combine this with the funnel and the email capture process for the site leavers mentioned above and you could well flip this into a profitable process.

    Like always, this just demonstrates that taking action teaches us something. Sitting still and doing nothing gets us nowhere 🙂

    Finally, the traffic situation is always an issue. Paid or free 🙂 For fast testing, paid is always a great option. Long term testing can make use of free as well.

    Facebook buy/sell pages are growing in popularity, so using these in the right manner, may well bring up some profitable campaigns, so I think I will give those a try and test out. Who knows, quizzes may turn into a nice income stream to play with to test other campaigns.

    Thanks again Thrive, you rock as always!

  • Great post Stephanie! One question: how would you approach a “non-physical product” niche, like – say – spirituality, writing and such? Thanks! 🙂

  • Great insight. I’m currently planning something very similar so this was really helpful.

    We sell our products to other businesses and we are going to use the quiz function to target maintenance staff in the engineering industry. The quiz will measure how effective their current maintenance strategy is and once a participant completes the quiz they will be re-directed to a unique results page with links to our products and services.

    I’m interested in adding a sign-up form (requesting name, email and phone number) as a call-to-action before participant’s take the quiz. Have you guys ever tried this before?

    I’m not sure if it will put people off taking the quiz but my aim is to keep the sign-up form simple. This would mean I’d have a list of data to import into our email marketing strategy (if they opt in).

    I’ve got my fingers crossed that it’s a success as we’d be able to replicate it across our other services and I can join you in the Bahamas sipping on cocktails 🙂

    • Hi Christopher

      It’s better to focus on one call to action at a time, so either them filling out the sign up form or taking the quiz. If you overwhelm them with more than one they’re more likely to just click off the site.

      I would recommend having the quiz first (great choice of quiz btw), then once they’re invested the time and want to know the result, add an opt-in gate before they can get their results. This can be required or optional, so you can either not show them the results until they’ve entered their details, or allow them to access their results either way.

  • That first commenter Martin seems to have some magic formulations and ways of doing things.

    My conversion rate is 9% or 8.98% to be exact. I buy no traffic at all. I get 300,000+ visits per month all organic from search, referral and social.

    All I do is text links to Amazon products, not even a review. 100% true story. Half the time people click on it to find out what something is or worth, then buy once they’re on Amazon.

  • An excellent experiment worth tweaking for a better result. And you’ve inspired me to do just that. Thank you, Steph!

  • {"email":"Email address invalid","url":"Website address invalid","required":"Required field missing"}