Thursday, March 22, 2007

Why most of the CPA/Brokerage industry will not be around in 5 years.

from Digital Moses March 22, 2007
by Adrian Bye

Many people currently involved in the CPA industry feel that this industry is rock solid and not likely to change anytime soon. They may be in for a shock. There are developments coming from technological, business and legal areas which are going to have huge ramifications on the industry. One of those just happened.

Specifically I am referring the brokerage fees that CPA networks and brokers charge (around 20%) to push offers to fill the internet demand for remnant inventory, and the inefficiencies and expenses that are put in place by having so many humans involved in making web advertising work. Over time these people will be replaced by technology, just as many industries have been overturned in the past by modernization causing blue collar workers to lose their jobs.

The three biggest sources of traffic for a typical network are:

a. Email marketing

b. Pay per click traffic

c. Web inventory such as banners and text links

We’ll talk about these one by one.

CPA search marketing

Pay per click marketing is changing. Google has just announced it is going to offer a CPA model. At the moment it is possible to make a decent living by being good at PPC arbitrage of CPA offers. This works right now because Google and Yahoo have focused their business model around selling clicks, rather than selling actions. They do this because this is their version of branded CPM advertising – they can generally get more for it.

However, this causes huge inefficiencies in the system, because it is time consuming and complicated to figure out how to drive lots of PPC traffic, enabling therefore arbitrage opportunities.

Since Google has now started offering a CPA system, and Yahoo certainly will, this will change dramatically. Advertisers will be able to add a bunch of creatives into the system, along with a list of keywords and a CPA they are willing to pay. The system will then automatically test the base keywords you inserted, along with an extra list of keywords google generated itself. It will test them all against the various creatives you made; keeping pricing under a certain CPA you have set. The entire system will be fully automated, and the current arbitrage which is possible today will go away. Google and Yahoo can make quite a lot of money by making this change, given the average network commissions and the money made by PPC arbitrage players. Google has already switched and it is just a matter of time before Yahoo does as well.

Notice I don’t mention clickfraud – I don’t believe this impacts Google and Yahoo moving to a CPA model.

Email marketing

Email marketing is an area which is going to change on two fronts. CANSPAM is a law with many loopholes, one that allows people to send as much mail as they want under certain (not very restrictive) limitations. A lot of mail is being sent which does not provide true value to consumers, its simply mass market monetization that is driving volume, a process I really disagree with. At some point a new email law will be passed which requires something like "at the time of sign up, the sending FROM address must be displayed clearly so the consumer knows where they will receive mail from". And brokering of email data will be exclusive only. You join one list, you unsubscribe from that list, period. It’s only a matter of time until something like this is legislated. Don’t think so? A few years ago the telemarketing industry was doing great – now its been decimated with the FTC’s do not call rules. This kind of thing can happen literally overnight – look at how the online gaming industry has been affected recently.

Secondly, deliverability is going to get much more difficult. Right now, most ESP’s can get mail delivered almost anywhere except major ISP’s such as Yahoo and Hotmail. Reputation management is a new trend in email which will change this dramatically. Reputation management means that every IP address which is sending email is certified by an independent third party as to how that IP address is being used to send mail. It provides a lot of data to email receivers on exactly how that IP address is being used. If you’re certified and your reputation is positive, a lot of your mail will automatically be delivered. If you’re not, you’ll get blocked as spam.

Right now reputation management is being used by the major ISPs to confirm mail delivery – but once this is rolled out more widely across internet mail servers, mail blocking will improve dramatically, and those who are sending bulk co-reg data will find their deliverability falling through the floor.

Behavioural targeting

Thirdly, behavioral targeting is going to get much better. This has been talked about in the past, and never seems to truly work properly but it is starting to get much better now. Networks like Blue Lithium and turn.com are making a lot of progress with targeting and collecting a lot of data on their userbase. Reports I hear about Blue Lithium in particular are that it performs extremely well.

Impressive things are being done on the advertiser side to take advantage of behavioural targeting. For example, Think Partnership has a new product called Second Bite which saves shopping cart abandoners. If you decided not to buy a product and half completed your shopping cart, Second Bite will work to get you to finish your purchase. Think Partnership is just starting to buy banner inventory to save the cart purchase. What this means is that you can be browsing the web and you’ll see a banner saying "hey – come back and finish your purchase on and get a 10% discount". Once this kind of inventory is brokered out to major behavioural networks, no general interest CPA offer will be able to compete with the CPM’s they will be able to pay to save a purchase. Sure, this is a narrow application, but many more clever targeting applications of behavioural targeting will appear, increasing CPM’s across the board.

In addition, client side behavioral targeting will increase. By this I mean that users will allow more data to be mined from their computers locally, and some of it will be passed back to the network. In an extreme case, imagine if Microsoft made its Windows OS completely free – but in return for being able to mine behvioural data from your machine. This data would be fed back to online targeting networks such as Blue Lithium, to target web advertising more accurately. No popups or any other nasty applications would be included. That could be a huge value add for consumers – with free software AND better advertising. Yes, this has huge privacy implications, but over time these will be worked out – the ECPM increase from accurate targeting will be too valuable, and consumers will not mind their data being mined in aggregate.

That’s not to say that everything is bad. Some areas of the CPA and brokerage industry will continue to work well. These include:

1) Coupon and affiliate sites. Publishers that are actively going out and finding links to promote on their site for consumers will continue to make money and want to use CPA networks. The human interface in this instance provides tremendous value to consumers since the publishers truly understand what their market wants.

2) Newsletters. This will become the standard for email marketing as the more aggressive forms of email marketing will be made illegal. This is similar to coupon and affiliate sites where publishers will actively seek out links to target their audience due to their understanding of their market.

3) Web and chat spam is going to increase. Right now we’re seeing quite a bit of spam on myspace, and given the progress people are making on defeating CAPTCHA mechanisms, this will only increase. If the postings cannot be effectively blocked by computers, then more of it will be done. Unfortunately CPA networks will see more volume from various forms of aggressive webspam as time goes on.

The branding industry will have some impact on these, but it likely won’t change much from the way it is now – some inventory will be sold at higher ECPM’s for major brands, and the rest will be remnant inventory. Of course the big question is how high the ECPM’s can get for behavioural targeting and whether they can beat branded advertising.

Some people will read this article and be thinking to themselves "no, he’s wrong, this has always worked, and it will continue to work". The reality is that the internet marketing industry has been around for a very short time, and we really don’t have any data points to compare against long term. The right way to think about it is "where is the true value for consumers and advertisers". If your business model doesn’t provide true value to all stakeholders, then at some point what you are doing will stop working.

If your business model depends entirely on brokering, you may want to consider how you will operate in a few years time once the above become reality.

A good way to think about whether your business will be around in the future is simply by answering two questions:

1) By running my business, am I creating true value for all my stakeholders (customers, employees, consumers, partners)

And

2) Am I keeping up with the very latest trends that might affect my business, including industries that are not directly related to my daily focus?

For number 2, you can say you’re doing the right thing because you’re reading this. J

Does this mean that all CPA advertising and lead generation will go away? Of course not. These are very fundamental models and the backbone of internet commerce.

Just watch out if your business model is entirely focused around brokering remnant advertising. If this is your primary business, make sure you stay on top of your strategy. You don’t want your company to be made irrelevant like has happened with generations of blue collar workers in the past.

No comments: