The Performance Marketing & Display Network Problem (Part II)
I wrote a few weeks ago about the issues with performance marketing and display networks, especially with the focus of cost-per-action and cost-per-lead campaigns. As mentioned, I’ve been involved in both the CPA affiliate marketing business and the interactive brand marketing business and know both of them inside out. Also, as mentioned I see the benefits of both, but see that performance based marketing is the future of our industry, and all agencies need to wake up and start learning about this part of the industry immediately.
One of the growing issues in performance based marketing when working with display networks is the often poor performance of display marketing when promoting CPA-type campaigns. Every major display network, while claiming that they do not allow CPA campaigns, has someone who works full time on accepting, optimizing and growing these campaigns. While they are often the fillers for the network, raising the ground-floor CPMs, more and more often they are high-CPM, high revenue generating advertising. However sometimes these campaigns work for the display networks but the quality is often horrible for the actual advertiser, even though they are paying on a CPA.
If you are curious how that is, please note that the vast majority of the CPA campaigns are not “cost per sale” type campaigns but are signups, lead generation or even brand campaigns with a CPA metric. Dating sites, for example, often pay per new sign up of users, and insurance companies often will pay for “lead interest” forms to be filled out. They hope through having an easy action they can often collect information, and then back out the consumer later for a sale or paid signup. However, as mentioned, often it doesn’t back out – and the advertiser can lose a lot of money quickly on a display network.
The reason often has to do with the lack of understanding how display marketing works and more importantly how it optimizes. In an email marketing campaign, for example, one is often reaching a broad range of consumers – all ages, all demographics. In the dating site example, all type of people will sign-up and the conversion rates from a trail member to a paid member usually hit a pretty regular metric. Since you are emailing a large group of people, you will get men, women, poor, rich – all sorts of people. However, in a display campaigns often something completely different happens: it optimizes very fast to show more and more advertising to the audience or site that converts the best.
This can be a significant issue because that means often only one demographic is often reached. In the case of the dating site, this demographic can often be young men who want to look at photos of “hot girls.” In one year I made millions on dating sites, because I designed ads that I knew would get guys to sign up.
Sites that have that audience in a display network will naturally as part of the optimization process show more and more of that advertiser’s ads. Other demographics will not be reached as much in proportion because they do not convert as much. Therefore the signups will often be significantly skewed and the 
quality extremely poor.
The solution to this is easy: advertisers need to understand how display optimization works, and take steps to produce higher CPA values for higher quality visitors with better demographics. If a network claims to have women over 50, and that is your target audience, reward the network with higher a higher CPA for signing up just that demographic. You will get other demographics but the high CPA will make up the difference in the discrepancy plus provide insight into that network’s real demographics.
I’m curious what other people think about this – and what agencies are experimenting with the CPA/CPL model to “prove” the results and demographics of their advertising campaigns. Love some feedback and insight.
Best,
?
3 Responses to “The Performance Marketing & Display Network Problem (Part II)”
Leave a Reply
(C) 2010 Pace Lattin. It is the policy of this blogger to not edit or remove any content and comments, unless it is specifically attacking a protected group or irrelevant to the conversation, such as a spam. These are the OPINIONS of the respective writers, posters, commentators. All DCMA Notices shoudl be sent to pace@pacelattin.com
Hi Pace,
I find your editorial to be insightful and although I really am outta my league when it comes to internet marketing on your scale, I find the information you provide very easy to digest.
If I read your article right, I think this is what we have tried to do with regards to http://dressupgames4you.com
Trying to create a media channel comprised of a very narrow demographic, giving the advertiser more credible target and lead acquisition .
Having redesigned the index to more effectively funnel our demographic to covert, we now see a marked increase in conversion to leads , up 17%
Having experimented in the cpa market for a very short time, I think that this is a great sign.
Not all demographic targets have money to spend online and are only an influence on the person who does make the purchase. Brand marketing and cpa marketing should still be one of the best ways to acquire quality awareness and potential customers.
Thanks Pace, keep it coming
Dale
Good Info. I totally agree with this post. Good Info Pace.
Ryan
Habitat Agency
http://www.habitatagency.com
Yes I agree. To counter the self optimsing nature within display towards the biggest clickers which only gets more of the same narrow demographic a different approach can be employed with the ad creative. Giving consumers a carosel of offers/creatives within the same banner will encourage a wider demographic to opt-in. For example in automotove don’t just display the family car, but let consumers scroll across every car model within the same ad giving a great chance of opt-in from every demographic and giving the display network a greater chance of a good eCPM.
[...] The Performance Marketing & Display Network Problem (Part II … [...]