A good question came to me in the email yesterday:
Is there a certain metric you use to predict revenue potential for a site?
I have a potential client with a blog that gets as many as 75k uniques and close to 200k pageviews per month… He can’t understand that he has tons of great content that people like reading and that it could be monetized.
There are indeed ways to put together a tentative projection of your future affiliate earnings. While the accuracy of these projections will always be highly contingent on a number of program-specific metrics (2 through 5 below), as well as on how exactly you are going to market merchants on your website, there is a number of variables to consider prior to starting out as an affiliate — metrics that will tell you how well your website may do with affiliate marketing. Consider the following five:
- Clickthrough Rate (CTR) — It will be dependent on (a) the targetedness of your traffic, and on (b) the type of linking you’ll be using (banners, text links, product links, etc). Read more about it here.
- EPC, or Earnings per 100 Clicks — This metric tells you what other affiliates are already making on this program, and while it should by no means be your single criteria (because it’s an average calculated across all affiliate activity within the program), it is a good one to analyze. More in my article here.
- Average Order Value (AOV) — Average ticket which is calculated as a sum total of all affiliate-generated orders referred over a given period of time, and then divided by the number of orders.
- Conversion Rate (CR) — The click-to-sale or click-to-lead conversion rate. Analyzing a program-specific CR is easier, especially when it is publicly available to affiliates prior to joining an affiliate program (as is the case with AvantLink-based programs, for example). Analysis of vertical-specific CR is also an option, but won’t be as precise. More on onversion rates here.
- Reversal Rate (RR) — This is the percentage of affiliate-referred transactions that get reversed by the merchant. Unfortunately, in many cases you won’t be able to analyze reversal rates prior to actually joining and experimenting with an affiliate program. Some networks, however (like AvantLink or ShareASale) do provide data on this vital metric even before you sign up with an affiliate program.
Plugging all of the above into the originally described scenario (publisher with some 200,000 pageviews a month), picking a merchant who pays 12% commission (comm) on all sales, presupposing a very close niche match between the affiliate and the merchant, and assuming the following:
- CTR = 1%
- EPC = $60
- AOV = $125
- CR = 5%
- RR = 10%
We arrive at the following calculation:
- Traffic: 200,000 * 1% (CTR) = 2,000
- Sales: 2,000 * 5% (CR) = 100
- Earnings: 100 * $125 (AOV) * 12% (comm) – 10% (RR) = $1,350
To arrive at your personal projected EPC you will want to divide Earnings by Traffic and multiply the result by 100. So:
- $1,350 / 2,000 * 100 = $67.50
Comparing our projected EPC figure to the program’s actual EPC ($60), we see that we are not that far off, and our overall projection must be sufficiently plausible.
Finally, a word of warning: remember that things will differ drastically from merchant to merchant. So, take things easy (especially when there are investments involved) not to fall victim to your own incorrect projections.
15 thoughts on “How to Project Affiliate Earnings: 5 Metrics and Formula”
That is quite a bit of thinking for not knowing much about his site or content.
I like the math as I am a math guy and you picked good numbers in my opinion. Good post
True. I have no idea what his site’s content is, but provided he finds good merchant matches (hence, the “presupposing a very close niche match between the affiliate and the merchant” part), this should be a good place place to start.
As for the math, I actually picked real numbers (rounding them up, of course) from a real affiliate program. The only thing I had to assume was the conversion (but this was easy to guess based on the niche this program is in, and the ranking it has). It was good to see my “projected EPC” come fairly close to the real EPC of this program.
Good stuff as always from you, Geno.
This is helpful for me as I get questions from time-to-time from site owners that don’t understand the potential for revenue that exists. While I don’t believe there are riches unless you have pageviews in the millions, I think it’s important to be able to give some kind of estimate for others, like the guy who asked the question above. In some cases, sites will jump at any extra revenues, even if it’s just a few thousand bucks.
Pingback: Web analytics and conversion essentials > Smart Insights Digital Marketing
Very good insights indeed! However, traffic and sales are useless in your calculation. What you just need is CR. EPC does not depend on impressions and CTR.
The comprehensive formula is: EPC = (1-RR)*(AOV*Commision*CR)*100
Antonio, you must be misreading the above post. It isn’t about how you calculate EPC, but rather about how/whether one may “predict revenue potential for a site” if they know how much traffic they get.
I didn’t. I’m just pointing out that in order to get the $67.50 projected EPC you calculated, neither the pageviews nor the CTR are needed!
They are necessary only when you want to calculate the overall potential revenue for a site as you too said.
…and the latter is precisely the (main) question we’re answering here.
I was wondering what do “sales” and “revenue” mean for a company which works as an affiliate network.
I would appreciate your answer, thanks.
I am not sure I understand the question, Mia. Namely, this part: “a company which works as an affiliate network.” Are you looking for a revenue formula for an affiliate network? 🙂
Geno – thank you for the concise article and methodical approach!
If a site had multiple affiliate partners would the results be cumulative?
For example, 5 different affiliate partners with the same assumptions you laid out. Would total earnings be $1,350 * 5 = $6,750?
Dennis, when you say “affiliate partners” do you mean the advertisers that the site would partner with (on affiliate marketing basis)?
Since the only thing that you can assume as a “static” metric is the CTR (even though it will still be just an assumption), while the rest of the above-mentioned metrics will vary depending on the advertiser, no, you do not want to just multiply the amount by 5, but calculate this for each individual advertiser and then add up.
That makes sense. Thank you for clarifying.