Adverts on Facebook – Maximizing click-thru rates to become a wealthy affiliate
Adverts, or advertisements, offer an opportunity for me to drive traffic to landing pages while slowly building up organic traffic from SEO. It would take forever for me to experiment if I had to wait for SEO to generate large enough sample sizes. I wouldn’t easily be able to determine if particular designs or adverts worked well, and I would fail while trying to become a wealthy affiliate 🙂
Currently, Facebook advertisements are some of the cheapest paid advertisements to drive traffic to websites. Additionally, the ability to target individuals with certain segmentation makes this a great platform to advertise consumer products like mattresses (the product of my mattress affiliate site). My previous blog post about the design of my affiliate site offers insight into how I designed my landing page to target mattress shoppers. Note: This landing page is not yet optimized…it’s just a tool to get started with affiliate marketing.
Optimizing adverts via variables – Background Information
Today is day 1 of my journey to become a wealthy affiliate. I’m going to start by defining what I mean by a successful advert. In my mind, the goal of online advertisements is to get viewers of my advertisement to purchase mattresses from Amazon (aka “conversions”) after visiting the MattressNut website. Here’s what a viewer needs to do in order to convert to a purchaser. Each of these steps is then a possible variable to optimize in order to maximize our advert’s efficacy:
Figure 1. Adverts initiate visitor engagement. Here is how visitors become shoppers.
Optimizing the customer flow means that I need to modify both my landing page as well as my advertisements. I’m going to focus on the advertisement first since that’s just a matter of a few clicks or images. For facebook adverts, I don’t need to write any new content or code in order to start experimentation.
Setting up my advert experiment – Experimental Setup
Based on the image above, there are a lot of variables to optimize. In my next post, I’m going to start by only testing the age of my audience. But, before I do any kind of optimization, I need a baseline. For my baseline I decided I would setup a quick ad, put some money behind it, and use my gut instinct to get *some* data back. I know I don’t have an ideal strategy or advertisement design, but I like to track progress vs. past data, and this allows me to get data quickly.
In this first “experiment” I’m just trying to get some baseline data with a quick and dirty advertisement on facebook. I’ll optimize my advertisements through detailed experiments in the future.
Here is the facebook advertisement and list of advert settings that I utilized (Figure 2 and Figure 3). Note that I didn’t run the Ad for the full 7 days, I stopped at 1000 impressions because that provides enough opportunity for the advert to run. Also, I know that I did not do any analysis of significance with this experiment. In future A/B tests I will provide p values for the data presented.
Figure 2. Facebook advert used in Experiment 1.
Figure 3. Setting used for the Facebook advert in Experiment 1
To track the traffic I utilized Taglogger, the tool that I created for user conversion tracking and discussed in my previous blog post about Targetted Ads to Shoppers vs. Visitors. Data was then cross-referenced vs. Google Analytics, Facebook Ad Manager, and Amazon to make sure that the results were consistent.
Taglogger was utilized to track user conversions and provided information that even Facebook didn’t provide me!
Am I a wealthy affiliate yet? – Results & Discussion
Facebook provided some really bogus results to my first advert. According to facebook results, the adverts were clicked 12 times (Figure 4). However, my Taglogger tool found that 119 visits to MattressNut came through the Facebook adverts (Figure 5). As I started noticing these odd results, I utilized an individualized source tag to track the visitors coming to MattressNut on GoogleAnalytics. This showed me, that of the 15 views of MattressNut, through Facebook, all the visitors bounced and spent 0 minutes on the website (Figure 6).
Figure 4. Advert results from Facebook Experiment 1.
Figure 5. Taglogger recorded these results from the Facebook advert above.
Figure 6. Google analytics provided interesting information about the traffic I received from the Facebook advert. Notice the avg. session duration on the site.
Clearly, this is odd behavior. Several questions pop into my mind:
- Could traffic from Facebook just be really poor quality?
- How can every single user that came over from my facebook advert bounce in 0 seconds?
- What are these bogus click-throughs to my affiliate site? And how is TagLogger measuring something so different compared to Facebook and Google Analytics?
- If this traffic is real, how do I drive up the time that visitors spend on MattressNut?
- If visitors aren’t interested in converting to mattress shoppers, would a different niche generate more affiliate revenue?
In Experiment 1, my facebook advert provided me very unusual results. There’s a lot of inconsistency between Facebook and Google Analytics/TagLogger. Several questions must now be answered in future experiments.
In my next post, I will start answering some of these questions that are brought about from this first baseline experiment.
Snapshot of my affiliate journey so far
So far I’ve only been able to setup my affiliate site and drive a small amount of traffic to it via Facebook Ads. However, as outline above, my ads have given some bogus results and I need to dig in more. Bellow you’ll see a snapshot of my earnings and expenditures so far (Figure 7). I’m no where close to being a wealthy affiliate yet 🙂
Figure 7. MattressNut Growth Year to Date (Nov).
Keep learning from my mistakes
I would love to connect! Follow me/Connect on LinkedIn and come join the OneStopMarketer blog.
Disclosure: I am a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.