Have you already checked out our guide on running experiments with the Geniuslink service? It’s a thorough tome of knowledge that’ll give you the how and why for running experiments with your affiliate links to ensure that your setup is performing at the highest level possible.
In that guide, we present two key examples for testing, and one of them is testing to see if Choice Pages perform better than direct-to-Amazon links.
We’ve been running a similar research project for the past five months. If you’re familiar with the encouraging results from Armando Ferreira’s case study, we’ve been seeing if we can consistently match those results. And, well… apologies for the tease, but we’re simply not ready to share these results just yet!
For now, we’re excited to share the testing process we developed and provide you with a guide to walk you through how to run this same test using your own links!
At a high level, the goal for this experiment is to measure the earnings-per-click for US traffic on links going directly to Amazon.com compared to those going to a Choice Page with two options—Amazon.com and B&H Photo Video.
Logistically, we’ll be taking your existing top links and converting them into Advanced links with geni.us links nested within. We then create two cloned links we’ll experiment with: one that goes directly to Amazon (the “control”) and another that leads to a Choice Page (the “test”). As we want to test only for US traffic, we use an Advanced link to send 50% of a link’s US traffic to a Choice Page (B&H + Amazon) and the rest directly to Amazon.com as normal. We’ll break this out, step by step, below.
We consider this to be a very low-risk test since only 50% of your US traffic is getting a different experience from the typical behavior.
Once enough traffic has passed through the amended links, we compare the earnings to see whether the direct link or the Choice Page performed better. We judge this based on the earnings-per-click of each Amazon ID + B&H.
Here’s an Outline:
1/ Generate two new Amazon affiliate tracking IDs for the test, and three Geniuslink groups.
2/ Identify which links to test with (5-15 links with significant traffic).
3/ Organize your Spreadsheet.
4/ Clone the original links twice: one becomes a Choice Page, another for the direct-to-Amazon link.
5/ Edit the original link so it splits US traffic into the two cloned links.
6/ Double check your work
7/ Wait for enough traffic through the links
8/ Run reports, analyze results, and do a happy dance.
9/ Wrapping Everything Up
Before we begin
In an attempt to reduce the variables, our own experiment involved creating Choice Pages that link to Amazon and B&H Photo. As a result, this guide is written with that goal in mind. You are welcome to use a storefront other than B&H—or even multiple retailers—in your own experiment.
Additional tools required: we recommend you use a spreadsheet for tracking your links, and an expanded clipboard app (which allows you to hold onto more than one string of text at a time) can definitely help.
1/ Set up the affiliate tracking IDs and groups in your Geniuslink account
1.2/ Generate two new Amazon Associate IDs in your US account. To do this, follow the quick walkthrough starting at 1:19 in our Groups & Overrides video. For example:
1.3/ Build three new groups in your Geniuslink Account
- “Choices Test” – Choice Pages
- “Choices Control” – Direct-to-Amazon links
- “Choices Track” – This exists to sanitize reporting
1.4/ Set up Affiliate Overrides for the Test and Control groups, using the new Amazon IDs.
- Choices Control – US override: glchpcontrol-20
- Choices Test – US override: glchptest-20
- Choices Track – default affiliate settings
2/ Identify the links to use for the test
2.1/ One option is to simply use the highest-trafficked links to ensure the experiment is completed as quickly as possible.
- Go to the Reports page.
- Set the date range to the previous 60-90 days.
- From the ‘All Links’ dropdown, select ‘Amazon links’.
- Set the viewable rows to 25
- Copy the list to your spreadsheet
- A 30-day report also works! If your content is updated regularly and you want your most recent links to be used for the experiment, running a 30-day report may be best
2.2/ Alternatively, some YouTubers opt to use the top 10-20 (or all) of the links in their ‘Gear I Use’ text block. Others may just use the links in their most recent videos.
3/ Organize your Spreadsheet
3.1 We’re going to be building two link clones for each link we’re testing, and it’s going to help a LOT to use a recognizable vanity code. Create two new columns and begin filling these in.
- For example, if your link is for a Sony a7iii camera, come up with two vanity codes that are identical except for the last letter, like so:
- geni.us/Sonya7iiicamCH (Choice Page)
- geni.us/Sonya7iiicamA (Amazon link)
3.2 With B&H and a whole host of other programs, our platform is able to find product matches for Amazon links with our Product Suggestions feature. But if you’re using a secondary storefront other than B&H—one that isn’t covered by our Suggestions list—then you may need to manually collect the product links:
- We got in the habit of opening up each Amazon link and copying the product title, adding it to the spreadsheet.
- From there, use the product title as a search query on your storefront of choice.
- Save these product links to your spreadsheet. We’ll need them later when building Choice Pages.
4/ Assemble new clone links
4.1/ Grab a link from your spreadsheet and locate it on the Geniuslink ‘Links’ tab. The fastest way to do this is by running a quick search using just the vanity code. (omit “https://geni.us/” from your search query)
4.2/ Clone the original link. Using the screenshot below as a guide, make the following updates:
a. Convert this clone into a Choice Page, if the product is automatically found on B&H via Suggestions, then you’re in luck! Just click to add it.
- If it isn’t, manually search for that product on B&H (or your alternate retailer of choice) and insert the URL as a secondary destination.
b. Assign the vanity URL from your spreadsheet (e.g. /sonya7iiicamCH)
c. Move it into the ‘Choices Test’ group.
- This ensures that clicks funneled through your Choice Page will be attributed to the glchoicetest-20 tracking ID
d. Leave the default Note in place: “Clone of geni.us/xxxxxxx”
e. We recommend changing the Call To Action text to ‘Find the Best Deal Below!’
f. Select the ‘Light and clean’ theme, which is recommended for physical products.
g. Click Save to generate your new Choice Page
4.3/ Once you’ve saved your newly cloned link, you’ll be looking at the top of the Links list with your new link highlighted in green. Next, we’ll clone your brand new link. This second clone will become the direct-to-Amazon link.
On the Link Creation screen:
a. Click the link type drop-down menu and change the link to a Simple link. The top-most button of the Choice Page will become the singular destination, so make sure your Amazon link is there.
b. Drop-in your second vanity URL. (e.g. “/sonya7iiicamA”)
c. Move your link into the ‘Choices Control’ group.
- This ensures that clicks funneled through your Choice Page will be attributed to your “glchoicecontrol-20” tracking ID
d. Click Save.
5/ Convert original existing links (from step 2) to Advanced Links
5.1/ Now we’re going to edit your original link so it redirects traffic into our test and control links. Starting with the top link on your spreadsheet list, open up the ‘Edit Link’ menu and convert the link into an Advanced Link.
Default: Amazon.com link (original destination, no changes)
5.2/ Set one rule: (If) (Country) (United States)
- Next, click the cog icon to turn this rule into an A/B Test. (How many of you knew that was there!?)
5.3/ Option A: Copy in the first clone link (Choice Page) 50% of traffic.
- This is the first place that those recognizable vanity URLs start to pay off!
5.4/ Option B: Copy in the second clone link (Amazon) 50% of traffic.
5.5/ Move the link into the ‘Choices Track’ group, again this sanitizes reporting by ensuring that non-US clicks aren’t counted with your control ID.
5.6/ Click ‘Update’ to save these changes. Repeat for all 5-15 links you’re testing
You’re just about done with the set up now!
6/ Double check a few links
6.1/ Go to the Groups tab and confirm the following:
- Do your Control and Test Groups have overrides?
- Do your Control, Test, and Track groups have the name number of links in them?
6.2/ Test one of the Advanced Links by copying it into the address bar. Where do you land?
For US-based clients, when you go straight to Amazon.com, look for the glchoicecontrol-20 ID as shown below:
If you end up at the Choice Page, click the Amazon button. Look for the glchoicetest-20 ID (below):
If you’re not in the US, use the /iso2/us testing parameter to observe US behavior.
7/ All done with setup!
Now, let those links soak for 30-90 days. Or until there are approximately 1,000 clicks reported into the ‘Test’ group.
8/ Analyze Results
Once enough time and traffic have elapsed, it’s time to see how everything performed! You’ll need to gather a few reports from Amazon and B&H. Then we’ll get real nerdy and pull some data out of Geniuslink’s dashboard. At that point, you plug all the data into our spreadsheet and we’ll have our answers!
8.1/ Snap a screenshot of your control tracking ID earnings during the time span of the testing period.
- Go to the reports section of the Amazon Associates dashboard, select your control tracking ID from the drop-down, and run a report starting on the day you set up the links, and ending 3 days before today. Make sure the clicks, commissions, and conversion % are visible. Grab a screenshot.
- Amazon’s reporting is only updated once a weekday and commissions can tend to lag. So we’ll exclude clicks and such from today, yesterday, and the day before. Take note of this time span, you’ll need it in a few spots.
8.2/ Repeat this report for the Amazon test tracking ID and grab a screenshot with the same data visible.
8.3/ Go to the B&H affiliate program dashboard and run a report for the same date range once again.
8.4/ Calculating EPC and reading the results
- Make yourself a copy of this spreadsheet. Within you’ll find instructions and formulae that’ll help summarize all your key metrics such as EPC, clicks reported by Geniuslink into each ID, approximate click-through rate, % difference between control and test links, and more.
9/ Wrapping Everything Up
Now that you’ve got your results, let’s put everything back into a clean state.
Would you like to keep all future Choice Page earnings flowing into your dedicated testing Amazon Associates ID? If so, leave the affiliate overrides intact. If you’d prefer all the earnings to flow back into your primary tracking ID, go ahead and remove the affiliate overrides from your Choices Test and Choices Control groups.
Next we’ll decide what we want to do with the test links themselves!
Option A: Keep the Choice Pages for all US traffic.
If you want to disable the A/B test, so that 100% of US traffic is sent to the Choice Page, follow this process for each of the advanced links in the ‘Choices Track’ group:
- Open up the Edit Link screen.
- On the rule for US traffic (the A/B test) click the cog icon and change that rule into a Choice Page.
- You will need to reconstruct your Choice Page in this rule.
- Ideally our Suggestions feature makes rebuilding the Choice Page a breeze. Just paste the Amazon.com link (from the default destination) into the field and we’re off to the races!
- Alternatively, open up the edit screen for your dedicated Choice Page clone in another window, and copy all the fields over.
- Click ‘Update’
Option B: Keep the Choice Pages for ALL traffic.
If you’re interested in presenting Choice Pages for all clicks, regardless of their country, then you’ll want to follow this process for all of your ‘Choices Track’ links:
- Open up the Edit Link screen.
- Using the Link Type drop-down menu, convert the link from an Advanced link to a Choice Page.
- Ideally the Choice Page builds itself with the Suggestions feature.
- If for some reason we’re not finding an automatic match in your secondary storefront, you can open up the edit screen for your dedicated Choice Page clone in another window, and copy all the fields over.
- Click ‘Update’
Option C: Leave everything as-is and continue the test!
Pretty straightforward! Perhaps you’re not sure you’ve got enough data, or you’d like to see how the test performs over a longer period of time. Since you’ve gotten this far, just record your data as normal in section 8 and let everything soak for longer.
Option D: Switch the links back to simple direct-to-Amazon links.
For each of the advanced links in the ‘Choices Track’ group, follow this process:
- Open up the Edit Link screen.
- Using the Link Type drop-down menu, convert the link from an Advanced link to a Simple link. All additional rules will be automatically trimmed out, and you should be left with a Simple link to Amazon.
- Click ‘Update’
Until Next Time
And that’s it! If you’ve read this far and followed along, at this point you will have some evidence-based data that should inform whether it’s worthwhile to adopt Choice Pages more broadly across your product links. If Choice Pages outperformed direct links, then this is that moment where you get to do a happy dance!
How did everything turn out? Do you feel like sharing your experience with us? We’d love to hear more about your test results or to be a part of your experiment in any capacity. Please reach out and let us know if you’ve got any questions!