Affiliate Marketing Websites & SEM - Exp. 5 - High vs. Low Daily Budget

Experiment 5: Is SEM viable for affiliate marketing websites? High vs. Low Daily Budget

Affiliate marketing sites should test their SEM ads
Affiliate marketing websites subsist on converting incoming traffic to their affiliate network. Utilizing the right incoming traffic sources is paramount to success. In case you haven’t already read what this blog is about, I suggest you learn more about how I’m trying to grow my affiliate revenue via a series of controlled experiments with my mattress affiliate website. So far, my experiments have included:

Out of these results, I have developed a few tools to aid myself and other affiliate marketing websites including Taglogger (a tool to measure conversion rates better than Google Analytics) and TimeLogger (a tool to measure user session durations better than Analytics).

Learn more about Advanced Google Analytics by reading this book about Breakthrough Google Analytics

Background Information: SEM Campaign optimization for Affiliate Marketing Websites

In my previous blog post about Google Ads vs. Bing Ads, I discussed how SEM may be utilized to grow traffic to an affiliate marketing website. However, a key part of this strategy is to scale up the traffic after initial experiments find the right combination of SEM ad platform + low cost keywords + useful traffic. Google Adwords definitely has the right platform + useful traffic, so I thought I could run a test of High vs. Low Daily budget while I continue to search for the right keywords (Figure 1).

How affiliate marketing websites can optimize SEM
Figure 1. Optimizing SEM based traffic and scaling up

Experimental Setup: SEM Scale-Up for Affiliate Marketing Websites

Setting up this experiment was simple since I was just comparing the same ad, with the same keywords, to the same audience (but with different daily budgets). I’m hoping that other affiliate marketing websites can also easily mimic this experiment before scaling up. Below you’ll see the ad I ran in AdWords (Figure 2).

Experiment ad for testing SEM for my affiliate marketing websites
Figure 2. Advertisement utilized to test the difference between high and low daily budgets.

Utilizing the advertisement above, I then tested a high daily budget ($50/day) vs a low budget ($5/day) for 1 day each.

Results & Discussion: Scaleup Results

The results from the experiment comparing $5/day vs. $50/day are shown in Figure 3. Notice that results from Google show that the CTR and Avg. CPC are equal even though the avg. position was slightly worse in the $5/day experimental group.

One might think this is a mundane experiment, but I believe it is necessary to test all my basic assumptions. I’m hoping that this information is valuable for other affiliate marketing websites that are testing their base assumptions.

Comparison of low vs. high daily budget SEM ads
Figure 3. AdWords results showcasing a comparison between $5/day vs. $50/day.

I also compared the conversion rates to Amazon from traffic that came through the two daily budget experimental groups. These results were gathered using TagLogger and are shown in Figure 4. It’s interesting that there is such a huge difference between the two daily budget groups. Note: the high daily budget was run on 1/19 and the low daily budget was run on 1/16-1/18.

Taglogger results for this experiment on affiliate marketing websites SEM
Figure 4. Conversion rates to Amazon from my affiliate marketing site from TagLogger data. High daily budget experiment ($50/day) was run on 1/19 and low daily budget ($5/day) was run from 1/16-1/18.

Why was there such a difference?

Upon close observation, I realized that the traffic from bots was skewing data.

How did I know bots were the issue? Google data shows fewer click-throughs to my ads than what TagLogger picked up. Taking out the visits from bots, I found that the conversion rates were very similar between the high and low budget experimental groups. This data is presented in Figure 5.

Without bots low vs. high daily budget has same results
Figure 5. Conversion rates to Amazon without bots (based on TagLogger data). High daily budget experiment ($50/day) was run on 1/19 and low daily budget ($5/day) was run from 1/16-1/18.

Weaknesses of this Experiment

It’s noteworthy that I compared 1/16-1/18 (multiple days) vs. 1/19 (one day) and that 1/19 was a Friday whereas the other days were normal weekdays. This could have skewed data but I was forced to try this out because of my limited budget for my business experiments.

Furthermore, going from $5 to $50 is only a minor change considering that companies may spend several thousands per day. Even though it is a 10x difference, there may be a larger effect on CTR and conversion rates once you go from $50/day to $500/day or even $5000/day. It’s possible that the kind of traffic that is sent to a website changes. However, this may be counter-balanced by keeping the CPC the same. It would be interesting to test how changing the CPC for the same keywords changes the kind of people who click-through. Is the intent of visitors from higher CPC traffic actually different than lower CPC traffic?

To learn more about experimentation with online marketing, read this book on Digital Marketing Strategy

Summary: What Other Affiliate Marketing Websites Can Learn

Based on this experiment, I found that changing from a $5 to a $50 daily budget does not change the type of traffic that visits my page, nor the click through rate. I also learned that a minor difference in position (3.3 to 3.5) has no major impact on traffic quality (because conversion rates were identical).

Follow my affiliate journey and marketing experiments

Sign up on OneStopMarketer or follow me on LinkedIn to continue learning from my online marketing experiments. It’s a lot cheaper for you to read my blogs, so why not get notified whenever a new blog post is put up?

Leave a Reply