Assessing an Affiliate Program

With the thousands of affiliate programs out there, tons of conflicting information, and endless strategies to implement, it can be tough nailing down what type affiliate program to join and what type of affiliate to market. The following is simple procedure we follow when determining whether or not to invest in an affiliate program…

Assessing Keyword Traffic Potential

There are a number of ways to do this find what way works best for you. [You don't think I would give away all my tactics did you?] I will however offer up something to think about when analyzing this data, no data is perfect, consider the source, and cross check if you are unsure. If you do everything perfectly and target the wrong phases it won’t matter if they rank, they won’t convert. Also, you’ll want to do the keyword research for the overall niche you target as well as any brands within that niche that will be on your site, this ensures you have a full frame to reference when making your decisions later on.

Determining Potential Average Conversion for an Affiliate

The idea here is to figure out what to expect from a given program before you put all the work into marketing it, this is only for estimate purposes, only once you get into the program and start making consistent conversions will you have the data to know what can be expected at true full potential. This is how I do it:

First off, I disregard the 3 Month and 7 Day EPC. I am not worried about what other people marketing the product have done, I’m not going to bring their poor performance data into my calculations.
Using the Commission Junction interface as an example, I click on the advertiser information link in the “get links” area in cj. This brings up the given advertisers program details, I go straight to the payout data. In the payout area you look for all the ways you can get paid by that advertiser, ensure you take into account any performance incentive(s).

You can see in the above screenshot the different ways you can get paid on the Verizon Broadband affiliate program. All you have to do is average the payout amount, so here I can expect an average conversion of $40, basically. However, it gets slightly more complicated if you have an established site and more data at your disposal. An example of this would be, if I know, from my keyword research earlier, more users are searching “Consumer FiOS Data” than “Business FiOS Data”, then based on that knowledge I can safely assume that the average dollar amount per conversion will go up.

Affiliate Link/Offer Evaluation

While all of these pieces are important, this can easily make or break it for us on whether or not we take on the program. Even with awesome traffic potential and legit payouts, if your affiliate program doesn’t offer the right links to the right page(s), I’m not going to make money easily. If the program is really good but has bad links, I just pass it up and will get around to it when I get around to it, which may end up being a long time. If you are an affiliate program manager and have ANY really big affiliates, that make 50k+ commission a month, aka a Platinum Affiliate, then you know one month without that high performing affiliate’s contribution can have a negative impact on your bottom line.

Give Affiliates what they Need/Want - Learn More

Again, using Verizon Broadband as an example, they actually denied our first application to the program, after I wrote them and sent them screen shots of our Verizon Wireless affiliate program performance, they apologized and pushed us an offer through CJ. September 2011 is/was our first month with the program and as of today 12SEP2011, we have already had 73 conversions, for a sale amount of $11,667.26, not to mention the fact that those are all contract based sales so they will be getting residuals off our leads for the next 2+ years. They could have been missing out on a lot of money, for no reason other than lack of attention to detail. Make sure you take the little extra time to setup links that are targeted and offer compelling value, this will drive conversions which motivates your affiliates to perform for you.

Affiliate Profit Potential Calculation

We’ll assume that I have just assessed Verizon Broadband affiliate program and it falls within our profile for investment verses potential. For demonstration purposes I will be using arbitrary numbers. This is how you calculate the potential:

Assessing Keyword Traffic Potential = 10000 searches per month for the various keywords I’d target for this program
Potential Average Conversion = $35 [We always estimate conservatively]
Link Assessment = Semi-Optimized, enough to proceed with campaign

Below is a visual representation of the percentage of traffic you can expect per ranking position in Google SERPs, or Search Engine Results Pages.

Percentage of Traffic by Rank in Google

Based on the above statistics, we do the following: 10000 x 0.4213 = 4213. The 4213 is the potential traffic we can capture ranked at number 1 across the board for the chosen target phrases. You now need to figure out what percentage you can assume will convert, if you have no experience in the niche you may want to be more conservative and go low to be safe. Across our network we average about an 11+% conversion rate, so we assume at least an 8% conversion rate when we do the calculation to error on safety. So of the 4213 captured, we convert 8% for a total of 337 conversions. If you take that and multiply it by the $35 per conversion estimate, we are looking at a profit potential for the Verizon Broadband affiliate program of $11795, per month. Now we decide if the target is worth the investment… To me, that number is a little low for a solitary target, but when you take into account how conservative that estimate is, it should make at least 2 times that amount, not to mention sites often have multiple affiliates targeted on them, the subject site that this would go on could do very well if I were to add a few other programs with this potential on it…

Of course getting your site to rank 1 in Google is the trick, isn’t it.