# Using EPC to make decisions

1. Using EPC to make decisions
I was recently posting here about using EPC (a merchant's average earnings per 100 clicks, averaged across all of their affiliates) as a gauge for selecting merchants. That post was about new merchants that are at \$0, but it opened an EPC thought train for me that I thought others may want to discuss.

What does EPC mean to you?

For me, EPC is just a very gross indicator of a program's health, like having a pulse. I contend it's not very useful for selecting merchants, or for forecasting your own likelihood to make a certain net profit. Let me explain a little further with some real, current numbers.

Rugman:
SAS overall EPC for last 30 days = \$74
SAS overall EPC for last 7 days = \$95
CJ overall EPC for 3 months = \$61
CJ overall EPC for last 7 days = \$44
My EPC at SAS for July month-to-date (as of 7/21) = \$153

I have a dedicated site for Rugman. I'm 2x the average EPC at SAS and roughly 3x the EPC at CJ (they have phone tracking at SAS for Rugman). I figure there are others that are considerably higher than me (couponers and offline advertisiers typically have higher EPCs for legit mathematical reasons), perhaps much higher than me. It wouldn't surprise me if there's an aff in the Rugman program, with plenty of traffic, whose EPC is well over \$400. If I turn down my ppc ads, I can drive my EPC well over \$600 and have done so at times (with much lower volume).

Other affs may just use a banner ad somewhere that's not as well targeted and converts at a much lower rate than the average (and if I'm 2x the average with lots of traffic, statisically, there must be lots of affs well on the other, lower side of the averages).

In other words, the deviation from the average not only depends on how well you do what you do, but also which method you choose to use and how hard you use it (generally, less pre-selling yields lower EPCs, while more pre-selling yields higher EPC and fewer clicks).

So given that we can expect EPC for Rugman to range, for affs with a reasonable amount of traffic, from perhaps \$20 to perhaps over \$400... just guessing, but I have seen and discussed the inside stats of enough other merchants to know I'm being conservative here... how does an affiliate considering Rugman as a partner using EPC? How do they give proper consideration to the range of likely EPCs?

Even the stats available in SAS and CJ range from \$44 to \$95...

If you try to project your revenue, traffic costs and resultant net profit, you'll get FAR different results depending on where you are in my "guessed" \$20-\$400 range, or even within the \$44-\$95 published averages range.

So I argue that using EPC, used as something to forecast your likely success with any degree of accuracy, is often going to be very inaccurate.

Compare your existing EPCs in your different programs against their current published averages and you'll see your own variations are likely significant.

Further, EPC doesn't account in any way for your cost of traffic. The calculation ignores those costs and assumes 100 clicks for one affiliate is the same as for any other. If you're an seo expert or email marketer and pay little ongoing costs for clicks, you're in a different league than a ppc affiliate with ongoing substantial costs per click. So in your pre-signup-with-a-merchant estimates, not only is their uncertainity in the reported average EPC, but (depending on your experience) there's also a lot of uncertainty in your traffic costs.

I like looking at EPCs (as a gross indicator) and am glad most networks make them available. But I'd say my review of their website, their AM's availability and track record, their current affiliate partners, their tracking network used, their popularity, their breadth and quality of products, their lack of leaks, their brand strength (people buy more often from people they know and trust) and other factors are all vastly more important than their reported EPC number.

If you're looking at two merchants that do/sell the same thing, some might argue that EPC is an accurate indicator of which will make you more money. I differ here as well. Since the average EPC depends strongly on the mix of affs and their chosen methods (which varies a lot from program to program, even in the same niche), I've also concluded through experience, that EPC is also not a very accurate indicator when comparing merchants who directly compete. If there were 3 being compared and they were \$8, \$43, \$60, I'd still look at all 3, but would be biased against the \$8 by looking "harder" to find leaks and problems that would explain their outlier EPC number... but I'd still look at them.

A quick glance at EPC can lead you to the wrong conclusions in many cases. That's my opinion anyhow, what's yours?

2. >>So I argue that using EPC, used as something to forecast your likely success with any degree of accuracy, is often going to be very inaccurate.

It can also be manipulated by an AM intentionally or unintentionally.

3. To use EPC to completely judge an affiliate program is to buy a \$300 pair of shoes because the box they come in is cool looking.
I think that EPC is a much more accurate guage of determining how you are performing in a given program once you have established a history.
If I have a rug site and my 90 day EPC after a year is \$30 and the published 90 day EPC is \$80, then I must think that I am not fully covering all of the marketing angles. If the opposite occurs, I must think that perhaps I am not using a broad enough brush to cover the program and am missing some business because my focus is too tight.

EPC can be affected by a devious AM or merchant. It is A factor, not THE factor.

I think more often than affiliates overemphasizing the importance of EPC is a merchant who kicks out affiliates that are not performing because they feel like these affiiates hurt their EPC. This merchant could be removing next year's top performer in the name of maintaining EPC and I see that as the merchant shooting themselves in the foot.

4. Originally Posted by UncleScooter
I think more often than affiliates overemphasizing the importance of EPC is a merchant who kicks out affiliates that are not performing because they feel like these affiiates hurt their EPC. This merchant could be removing next year's top performer in the name of maintaining EPC and I see that as the merchant shooting themselves in the foot.
Good point Scoot! I can deliver much higher EPCs for a merchant by turning down my input and by adjusting my amount of pre-selling. So less volume, much higher EPC. My ROI goes up, but my overall volume goes down... and the merchant loses sales volume. I agree, it's an anti-indicator for merchants. They want reach and aff marketing gives it to them (where they only pay for sales, not all that reach). So cutting off low EPC'ers is a backwards decision on the AMs part.

5. Originally Posted by Adam Ward
It can also be manipulated by an AM intentionally or unintentionally.
Totally agree here too! And the unitentionally comment hits home with me. Have seen some merchants partner with certain marketing partners where their EPC dives way down, but overall, the partner is a smart choice for the merchant. If affiliates give too much credence to EPC, AM starts thinking of ways to lessen "negative" impact of EPC and finds ways to take that partner off network through private offers / private tracking. Sometimes this is done with a cookie sensing mechanism much like dual platform. So what does average EPC come to mean, when some types are removed from the average? Or in an indie when the merchant routes their own ppc through the aff tracking system?

Affiliates should be aware that there's many things that can't be seen, legit or otherwise, that can drive published EPC in many directions - both intentionally and not.

6. Nice analysis Donuts. I'd like to also add that certain affiliates can affect a merchant's EPC.

You can have a well performing program, where many affiliates enjoy a high EPC, and have some affiliates dump huge amounts of useless traffic on you and drag your EPC down. If you're not watching your program closely, such affiliates can drag your EPC dramatically. If left unchecked, it can tank your EPC.

With Freelance Work Exchange, we get a good amount of traffic each day and our 30 Day EPC has been steadily growing. Our 7 Day EPC fluctuates wildly at times. It can build up over a matter of weeks and one of our partners can put a link in some traffic scheme and dump 5-6K clicks in a single day with no sales. I usually catch them fairly quickly, and have them remove us from such programs, but by the time it stops the damage is done.

I agree it's a tool and can be one of a number of useful tools to help judge an affiliate program. Someone "shopping" by EPC alone could be missing a lot. EPC doesn't take into effect many variables like how many affiliates are in the program and the quality of them. The top partner with Freelance Work Exchange has an EPC over \$97 this year, with thousands of clicks, yet our 30 day is a little under \$13. Their conversion rate is over 9% to boot. This tells me they certainly know how to deliver tons of targeted, qualified traffic. But, I would never tell potential partners that this is the "norm", possible yes, but not the norm.

Like Donuts said, your EPC depends on many factors from the merchant, leaks, program manager, products, your ability to pre-sell or how well you create your PPC ads. The big "unknown" to affiliates is just why a program's EPC could be very low (or very high for that matter). Do your due diligence. It's a business, treat it as such.

7. Originally Posted by Greg Rice
You can have a well performing program, where many affiliates enjoy a high EPC, and have some affiliates dump huge amounts of useless traffic on you and drag your EPC down. If you're not watching your program closely, such affiliates can drag your EPC dramatically. If left unchecked, it can tank your EPC.
Excellent point here too. Driving junk traffic to an affiliate site wouldn't hurt the EPC cuz they're unlikely to click through to the merchant. But if some dopey affiliate signs up with LookDog (or some other bullcrap autosurf program) and doesn't use an aff site (but shows the merchant's site using an aff link), EPC goes down the drain.

I have heard of times where merchants will do this to one another... sign up in your competitiors program and spam it to death with autosurfs and similar techniques to tank their EPC... so it's not just manipulation in-house that can occur...

Direct-to-merchant affiliates can be great affs to have on your team, but they directly impact your EPC by their quality. Smart AMs use networks that show you the referral urls (ShareASale, MyAp with Referral Tracking Module) so that can keep an eye on where their Direct-to-merchant folks are sourcing traffic.

8. Originally Posted by Donuts
So cutting off low EPC'ers is a backwards decision on the AMs part.
I agree. I wouldn't cut off an affiliate because their EPC was low, it's a short sighted move. AMs need to consider their purpose and their goals. If your goal is to have a solid program, is a win-win for all and makes money for your client then why cut off a low EPC partner?

AMs thinking about this type of move, consider this: Try investing some time by helping your partners sell more/better. A small percentage of partners actually produce each month. What if you worked with the under performers to get even one lead or sale per month? You're helping them be more successful, which could grow, and think about how this can affect your monthly performance. What if, say 20%, of your partners had at least one sale each month? What about 30%?

9. Further, EPC doesn't account in any way for your cost of traffic. The calculation ignores those costs and assumes 100 clicks for one affiliate is the same as for any other. If you're an seo expert or email marketer and pay little ongoing costs for clicks, you're in a different league than a ppc affiliate with ongoing substantial costs per click. So in your pre-signup-with-a-merchant estimates, not only is their uncertainity in the reported average EPC, but (depending on your experience) there's also a lot of uncertainty in your traffic costs.
This is a good lesson for you newer aff marketers. Don't look at this figure when you're calculating what you bid on keywords. Only your own performance can determine that. What percentage of your ad clickers end up at the merchant? Not 100% unless you're Donuts

EPC is also affected by commission percentage and genre of product, especially when you calculate in junk traffic.

If someone is selling \$1000 software, maybe YOUR site can generate 1 sale in a few hundred clicks, making your EPC exceptional. But junk traffic is even LESS likely to convert in this category.

Everybody can accidentally sell a pair of shoes once in a while. The question is who can sell the bigger ticket items...

10. Great thread. Learned a good bit from this. I've realized the hard way that it's nearly impossible to judge performance by looking at EPC. I've had high EPC merchants do squat for me, while low ones often times bring in the dough. Go figure..

11. Here's a crazy extension to my intial post... Let's say Rugman called me right now and doubled my commission percentage. I'd go out and search up bunches of more traffic. In all kinds of places. I'd certainly go after more keywords with higher bids in ppc and take out more print ads in papers, maybe nationally instead of just locally. Now, I base my business on ROI and volume. If my ROI is 200%, but profit is \$2 a day, who cares. I'd take 100% ROI for an average \$100 a day profit for instance. It's a sliding scale. If Rugman doubled my comms, I'd push beyond where I am now to the point of accepting a smaller ROI % in favor of a larger volume.

My earnings would go up cuz of the raise, but my clicks would go up more (on percentage basis) as a chased things further into the woods.

And guess what... my EPC would go down by a fairly large amount... so would Rugman's, but just a tad cuz I'm one of many affs in their aggregate average...

Maximizing EPC isn't a goal for me or for most of my merchants. It's just another indicator to look at - and its value is pretty low in the grand scheme of things for me.

Edit: By the way, neither Rugman or I would mind the lowered EPC cuz we'd have a lot more sales coming in the front door.

12. On a similar note then Donuts, I'm curious how you read your own EPC, never mind the merchants.

When you look at the performance of one of your sites, does YOUR epc determine what merchants stay and which go? Or do you have another way of calculating which of your merchants stays (aside of course from cold hard cash).

I know Rugman is a bad example here, becuase they are both unique in and of themselves, as well as you mentioned they were exclusive on your site...

Also, do you go so far as to calculate your EPC down to PPC EPC (if the ads go to your site first... if they go DTM, of course you would) versus EPC for organic traffic?

13. Something that I have been noticing lately with the EPCs and wondering whether they are a good way to gauge using particular programs is their "leak" factor and tracking accuracy. Is it fair to say that programs with a consistently very low EPC but that show a good deal of traffic are either not reporting all of the sales accurately or maybe leaking out some of their business through contextual advertising on their sites or sister programs?

Or is that presupposing too much?

14. While it's possible, the simple truth remains that some merchants just don't a very good job selling on their site. In fact, it's one of the factors that a lot of affiliates neglect to look at when signing up for a program.

Maybe a topic for a different thread, but how many of your merchants do you go through their website carefully and often?

But back on topic, many things can contribute to low EPC. Your mentions are certainly plausible.

15. Donuts, I agree that EPC is only one element in evaluating merchants. For my established programs, I never worry about it. It's the bottom line that counts. The total amount of money I'm getting out of a program. If EPC is low, but total return is high, that's fine with me. Maybe not with the merchant, though.

Deciding on adding a new merchant, however, I give a little more signficance to EPC. I'm dealing with something not completely unknown, but still a little bit out of focus. The EPC gives me some feel for how a collection of existing affiliates are doing. I have to assume that the affiliate base is made up of all kinds from low performing to high performing. The average EPC tells me how the whole group of them are doing.

I'll almost always come here to abestweb to see who is active with the merchant, who is complaining, and who is complementing. Then I plug the data into my scientific formula to make a decision.

Here it is:

Decision = {EPC + ½ω(σ*E8)φ[(SubjectiveThoughts)+(CoolProducts)*GoodAffManager}

16. Originally Posted by raywood1
I'll almost always come here to abestweb to see who is active with the merchant, who is complaining, and who is complementing. Then I plug the data into my scientific formula to make a decision.

Here it is:

Decision = {EPC + ½ω(σ*E8)φ[(SubjectiveThoughts)+(CoolProducts)*GoodAffManager}
Do you mind if I steal your formula but add my own "Subjective Thoughts" and divide the whole thing by the number of typos that I see on the merchant's website? (my own personal pet peeve)

17. Decision = {EPC + ½ω(σ*E8)φ[(SubjectiveThoughts)+(CoolProducts)*GoodAffManager }
I think this makes my question above look silly... LOL.

Is subjectivethoughts a constant?

18. Originally Posted by Noth
On a similar note then Donuts, I'm curious how you read your own EPC, never mind the merchants.

When you look at the performance of one of your sites, does YOUR epc determine what merchants stay and which go? Or do you have another way of calculating which of your merchants stays (aside of course from cold hard cash).

I know Rugman is a bad example here, becuase they are both unique in and of themselves, as well as you mentioned they were exclusive on your site...

Also, do you go so far as to calculate your EPC down to PPC EPC (if the ads go to your site first... if they go DTM, of course you would) versus EPC for organic traffic?
True to form, I don't track my own EPC stats at all. Once in a while, while looking at sales, comms and stuff, I see my own EPC stats as calculated within SAS for instance. But EPC isn't in my "books" I keep.

For performance, I track all my sites and programs on a daily ROI basis, both total profit and % ROI. Could care less about EPC once I'm in a program and, in the light you suggested, looking at my own EPC or my EPCs across different programs / sites. Since traffic costs varying amounts in varying places, I do often segment out traffic sources and sales intraprogram, but again not to grab an EPC, but to view the return, both aggregate and %, that I am getting for my investment. For me, in addition to cost, as a one-man show, I need to figure efficiency into all things done - if a traffic source requires a lot of time, but is cheap in dollars, I might dismiss it in favor of several others that take less time. Time is often more precious than dollars with the amount of work I have on my plate. Yet another reason why EPC doesn't matter to me - it's time blind.

Where I have multiple merchants on a site, it depends on how I have them there - if it's a side by side comparison, I might keep a lesser performers because it IS a comparison site - means more than 1 merchants are needed. If it's not a comparison situation, they're usually complimentary, not taking from each other - maybe I plant them on different pages / sections - so why remove them (if they're not really competing). My breaks with merchants usually come from gross changes they make to their program that throw the balance of things out if whack. I can tune for performance over many small bumps, but when a canyon appears in the road, and the merchant caused or allowed it to happen, I usually find another road to take. Long term reltionships work better for me - less maintenance and less time. Most merchants I'm with know me very well and we stay together for years and years, through thick and thin. Reliability and trust get me farther than pure short term return percentages. This also makes me a real pain to recruit - turtle like approach to enter, and even slower to depart.

And yes, I do track ppc and organic traffic seperately, usually via linking tag mechanisms, but there are other ways. My seo methods are small spend angles (pr, backlinks, text ads, etc) compared to my ppc - so for me it's critical to split out organic to know my true return on ppc (where I spend tons). If I allow seo sales to mask my ppc's true roi, I'm not able to work towards maximizing my ppc spend / return.

I do slack off on the weekends... :-) But Monday morning's work always includes grabbing Friday, Sat and Sun stats and entering / getting them all in my system for tracking what interests me (won't find EPC there). I do doodle with my numbers sometimes, looking for meaning in other views, but nothing beats max net profit, followed closely by max % ROI. The former is where I'm tuning my program's performance and the latter is more where I try to figure out where to spend my marginal time and money going forward.

19. Originally Posted by helpingmoms
Something that I have been noticing lately with the EPCs and wondering whether they are a good way to gauge using particular programs is their "leak" factor and tracking accuracy. Is it fair to say that programs with a consistently very low EPC but that show a good deal of traffic are either not reporting all of the sales accurately or maybe leaking out some of their business through contextual advertising on their sites or sister programs?

Or is that presupposing too much?
Leaks are the first things I look at, if I spot a few, I'm way gone.

One way to avoid all those troubles is to follow around certain Affiliate Managers... they know what they're doing and will plug the leaks. Andy knows if he calls me, I'll see leaks in 10 seconds... he usually starts telling me straight up about leaks before I see them and tells me the ETA for when they're already scheduled to be plugged. Michael Nunez is another I respect for knowing exactly what to look for - super glad to hear he's going OPM - he lives nearby and I've talked to him for hours and hours and hours - dude knows his stuff and like Andy, understands they need not waste my time ever with a call or IM about a merchant with leaks that aren't in the process of plugging them. Chuck Hamrick and Greg Rice are smart like these guys too.

I have been shifting my attention more to good OPMs and folks like Chuck (inside AMs) where I know there's not nearly as much for me to worry about when I jump into something.

20. Originally Posted by raywood1
Decision = {EPC + ½ω(σ*E8)φ[(SubjectiveThoughts)+(CoolProducts)*GoodAffManager}
Suggest you cube that "GoodAffManager" term in the equation...

21. Nunez's new program...

Skim the thread above... here's what I see. Affiliates point out problems at ABW. Nunez acknowledges them, without getting defensive. Says what and when he's doing about them. Posts when fixed and asks for more input. Nunez is out in the open, killing hurdles, clearing the way for affiliates to succeed. I'm very interested in his program and could care less what their EPC is.

22. Thank you for the detailed reply Donuts. I've bookmarked it.

23. I use EPC to judge my own performance with a merchant because it takes everything into consideration. It's the best measurement out there.

24. Great points! I made some similar ones earlier this year, along with a suggestion for an algorithm for merchants/networks to calculate a a range of EPC numbers rather than a simple average.

25. Do you mind if I steal your formula but add my own "Subjective Thoughts" and divide the whole thing by the number of typos that I see on the merchant's website? (my own personal pet peeve)
Everybody is welcome to this precise formula. It is based in part upon Flegal's Varible Constant, which I use for the SubjectiveThoughts parameter.

Donuts is the one who brought this topic up. He deserves credit for pointing out the fallacy of getting all spun up about EPC. It is a very good nudge to both newbies and old hands to remember that there are many other indicators and predictors of success with many variables for each individual aff marketer.

Too bad Micheals's ideas have been ignored. If implemented, they might make EPC a little more important.

#### Posting Permissions

• You may not post new threads
• You may not post replies
• You may not post attachments
• You may not edit your posts
•