Wow - James, this is one of your best posts. Alot of the advice for newbies is to look for high EPC advertisers on networks. I had never considered that EPC may be as much a factor of how picky the advetiser was - as in how well traffic converted. Also, CJ does seem to put more emphasis on recruiting new affiliates than retaining old ones. The old adage - it's easier to keep an old customer (or affiliate I guess) than get a new one - doesn't seem to apply there.

HERE HERE James! That has been the most frustrating point for me through the entire CJ management. I have let CJ know on several occassions that this is a large frustration. Although the account managers are empathetic, I think their hands are tied. I've suggested on a regular basis that this be improved somehow but that seems to go unheard. Thank you for the post.

Very interesting post and nice seeing a merchant or am saying how they feel about something CJ is having them do. :) One thought to something you said:
"if InsureMe is forced to expire a publisher because of poor EPC, that publisher may find and join our in-house affiliate program."

That publisher/affiliate will most likely only do that if the merchant or am takes the time to explain to them why exactly they are ending that relationship with them. If they are not told why it is happening....believe me there will be no warm and fuzzy feelings towards that am or merchant because they won't know that it was done without the merchant or am wanting to do it. Publishers/affiliates for the most part I think have elephant memories and are not happy when they have been terminated or expired) (psst... ending your relationship sounds so much nicer lol ) but an explanation and a willingness by the am or merchant to want to maintain a relationship with them will inspire them to join your in-house program in many cases.

Hey Quoteus – glad you enjoyed the post. I am a strong believer in building lasting relationships with affiliates…and I’m afraid CJ can make that hard to do. When an intermediary doesn’t allow you to access contact information easily, it really reduces the ability to for both merchant and affiliate to build anything lasting.

Thanks for the comment April – there seems to be a lot of rope for tying hands in the CJ office…hopefully they’ll wake up before they accidentally hang themselves with it.

Good point Kathy. I’d agree that communicating with good affiliates with low EPC and letting them know why you are expiring them is important. I’ll make sure I do that whenever possible.

Did I say expire? I mean “end the relationship.” :)

Interesting post. Responded at BumpZee discussion.

Make sure you check out the comments on this post going on over at Bumpzee by clicking here.

As a publisher I very quickly realised that EPC is very shaky and unreliable number.

Especially when you consider that often bot clicks will be counted in the EPC calculation, which further muddies the water.

I agree with you, it's a reall shame that it's all advertisers have to go on.

Perhaps, one of the solutions CJ could adopt when calculating EPCs, is a top & tail effect by not including the extreme figures towards both the top & bottom end of individual affiliate EPCs to establish a more realistic figure. This could be anything from 10% to 20%. Or maybe looking at the mean EPC. Or analysing the the top 20% percentile to provide a figure of what EPC is acheivable. Some of the above may prevent affiliates being rejected from programs who are generating sales but with a lower EPC.

I would like CJ to bring back trend graphs for EPCs illustrating two graphs. One showing a daily scale & another a rolling 1000 click scale along one of the x-axis, which if a trend falls unusually below a recognised threshold, it may help identify problems like lost tracking.

"The truth is different publishers send different types of traffic in different way"

Very valid point. Potentially there could be different EPC's based on those derived from brand bidding, direct to merchant from recognised ppc search engines & those from registered affiliate sites using product feeds or normal directory listings (again adopting any of the following effects - top & tail, mean, percentile). All in all providing greater data for affiliates & advertisers to make their own inferences from.

All in all I am in favour of EPCs being published, and experienced affiliates will know how to interpret these. However some information is better than no information.

The data is readily available for networks to provide more meaningful EPCs varying on the forms of traffic being generated. Maybe we should be grateful that at least some networks provide this information, whilst others are reluctant to publish these and/or merchant contact details.

Wow Moose...great ideas and a well though out comment. I know that CJ is aware of this post, and has been reading comments, so we can only hope they are proactively working to find more accurate/granular ways to display EPC information, just as you describe.

James, as a CJ advertiser I think this is a great post. I've also pondered the importance of the EPC .

There are other examples of not so productive ideas within the program. In example, the way CJ organizes top performing banners limits the profitability of others. To illustrate, top performing banners, rated by EPC, appear on the top of our banner list, while others, with low EPC, stay near the bottom never to be plucked out by potential publishers. Affiliates, be sure to fully explore your options before selecting a banner add for your site. Some of the newest and most innovative ideas are hiding at the bottom of banner lists. Bottom listed banners don't appear to have made as much money as banners listed on top, but it doesn't mean they have less potential to make money for you. Try to give them a solid chance

Visit AIG Travel Guard in Commission Junction. Search under Travel/Accessories. We are currently offering a $16 fee for every sale originating from a banner add or text link on your site!

Thanks again, James.

Matt