Les éditeurs doivent-ils avoir peur du CPA?

Suite à l’annonce de Google qui se lance dans l’expérimentation du modèle de CPA (Cost per Action) sur son moteur, Fred Cavazza a relancé le débat sur la rivalité des différents modèles. J’ai déjà écrit ici pourquoi le CPA constitue un piège pour les éditeurs en ligne et je constate à la lecture des commentaires que depuis un an les arguments encourageants à la méfiance se sont développés. Plutôt que de relancer le débat avec les mêmes interlocuteurs j’en ai profité pour poser la question franchement sur Linkedin (où figure mon CV) auprès de mon réseau de contacts. une bonne occasion de tester le nouveau service de questions/réponses du réseau Linkedin.

Je dois dire que le résultat dépasse mes espérances et j’ai reçu en quelques jours plusieurs réponses très argumentées de la part de professionnels du marketing en ligne de tous les coins du monde (USA, Allemagne, Argentine …). Un point pour les réseaux communautaires professionnels, cela confirme au passage la bonne qualité du réseau Linkedin. Je reproduis ci-dessous les principales réponses:

La question :  Should online publishers and medias be afraid of CPA (Cost per action) model ?

– Rick Monihan ( ex directeur Commercial chez Fox News-USA):
The CPA model is not one with publishers will ever be happy with, unless they have a direct interest in the acquisition portion of the deal. If their sole source of revenue is the money generated by CPA, then it’s always going to be a losing proposition.
Here is why:
1. Most advertising is lead generation – oriented. That is, an ad is generally designed to spread information and get consumers interested. It isn’t always a matter of getting an immediate sale or purchase. So, alot of ads sold as CPA will run without generating revenue.
2. Publishers are at the mercy of the advertisers. If the advertisers have lousy products, lousy ads, or a lousy website, then getting the Action or Acquisition portion of the advertising may never occur. More ads with empty revenue potential.
3. Brand equity…even with a good CPA campaign, there is a value to building the brand. So, with the CPA model the advertiser only pays for the A (action or acquisition) but doesn’t pay for all the additional benefits of advertising – spreading the information about the product, improving brand image, etc. That is all given for free by the publisher.
The solution is for the publisher to be directly linked in with the revenue generation portion of the relationship….but then this will erode the independence of many outlets that rely on advertising for revenue.
CPA can be PART of an overall revenue generation strategy by publishers, but it can’t be the sole revenue source, nor should it be the primary one.

– Eli Feldblum (Expert SEO, online marketing, Israël)
I say: Yes, very afraid.
CPA uses too many variable that you, as a publisher, can’t control. You can do everything right to your site: optimize it, bring in qualified traffic, keep it updated. But if your advertiser’s site has problems or doesn’t offer a compelling product, you don’t get paid for your valuable advertising space no matter how many impressions or clicks the ad gets.If you do want to accept CPA ads, go into it the same way you’d go into a business partnership. Accept ads on a one-by-one basis; if you would partner or revenue share with this site–if it’s compelling enough to make you commit whatever action they are paying your site for–then it’s good enough to accept on a CPA basis. If not, then no way.

Quelques positions plus originales et moins hostiles au CPA:  

– Fernando Garcia Penella (Manager chez Avaya, Argentine) 
Let’s take it from the customer point of view. A sophisticated online advertising customer will be measuring publishers and online media from a CPA point of view if it’s looking for a lead-generation campaign. And choosing the best value for its advertising budget.
Online advertising gives him the ability to measure this marketing expense and finding out which sites, portals, networks, etc. give him the best bang for his buck. And that is how he will evaluate in the long run, no matter whether he is purchasing by CPM, CPA, or whatever other metric.
So the online publishers are already being measured through this optic by this customer — they are already part of the customer’s business; the media that will offer him a CPA model is only speaking in the same language that this customer.
Obviously this requires a change of worldwiew for online publishers. It’s no longer advertising buisiness as usual…..

– Pierre Alexandre (Patron de New York French Press – USA -mais français expat…)
The advertiser want to pay for result. The one who sell expensive space withtout trust real audience need to be afraid. But those who have a good and REAL audience don’t haver to be afraid. Now the content is king. And the time of liars poker with advertiser is over.

– Ryan Turner (Consultant en communication – USA)
Publishers and media have nothing to fear if they’re being honest about the reach and loyalty of their audience, and not trying to cheat the system, regardless of any approach in place. By going the CPA approach, online marketing– especially in areas where risk and open-minds should be greatest– can turn into a « bounty for hire » that removes the sense of personal cost and investment in the effort’s success.Other legitimate concerns:
— brand separation (more emphasis on the ad, product, or overall brand at the expense of marketers pushing online transactions);
— brand dilution (mishandling of the brand through disconnect in marketers);
— prospect cannibalization (online channels and sites marketing at cross purposes with one another)At this point, remember Google is only testing CPA against CPC models, though I suspect in the short- to mid-term they will implement both as a hedge against competing providers, and offer different tiers or packages of services. It’s their best move to reach solve a combination of problems and hedge their bets in the short- to mid-range. Then they will eventually phase in one hybrid model that takes the « best » of both for them.

Think of it this way: of the following, which probably *make* more money for Google and which ones probably *cost* Google more money under a CPC scheme?:
— small-size/high-traffic sites
— small-size/low-traffic sites
— large-size/high-traffic sites
— large-size,/low-traffic sites

I’m guessing the small-size/high-traffic and large-size/low-traffic sites would the type of groups Google wants to court for a CPA program, because they’d wind up paying less than under a CPC plan while gaining enough of a base to make it worth their time, no?

Retrouvez toutes les réponses ici. Pour ma part je partage la plupart des critiques faites au CPA, je reste très surpris que certains y voient une opportunité de monétiser l’inventaire de pub invendu. Ceci alors que comme l’ont relévé plusieurs commentateurs, on voit de développer des pratiques douteuses consistant, côté annonceur, à acheter des campagnes de branding (notoriété) sur les réseaux d’affiliation afin de bénéficier de coûts de diffusion écrasés tout ça sur le dos de l’éditeur.

En fait je crois que le malentendu est là: il y a de la place pour le CPA chez les éditeurs simplement ce n’est pas une affaire de bannière et d’affichage. Le CPA est adapté sur les zones marchandes ou dans un contexte de shopping et de consommation et sur des formats différents (text, liens contextuels etc…) . C’est fait pour et comme le disent finement les derniers commentaires, si un éditeur fait la démonstration que ses lecteurs sont aussi des consommateurs il gagne en crédibilité et tire son épingle du jeu.