E-COMMERCE Knock-down, drag-out CPM vs. CPA (991 words)
For several years, the market for online advertising has boomed, even as it struggles to find a stable pricing structure.
Initially, banner and e-mail pricing resembled a WWF free-for-all with traffic juggernauts dictating terms to hungry marketers. When the recent technology bubble burst, it brought sanity to stock valuations, but added increased instability in the online advertising industry.
Fueled by a deluge of excess inventory, low barriers to entry and a slow down in the U.S. economy, many advertisers are attempting to fundamentally alter the rates and manner in which they pay for advertising by striking performance-based deals.
These cost-per-action (CPA) deals mean a Web publisher only gets paid when the advertiser registers a sale instead of the traditional cost-per-thousand (CPM) pricing structure where the publisher gets paid based on the volume of advertising delivered.
The acceleration of CPA deals is due to advertisers' attempts to rein in customer acquisition costs during these tight times, combined with publishers who are hungry for revenue. Both publishers and marketers have valid complaints about pricing. Advertisers argue they should not have to pay for low-quality placements, low click-throughs on banners and e-mail lists that are of questionable quality.
Sales, they continue, are the actual barometer of worth, and "a good ad spot or list should convert"—thus the publisher will be paid.
Publishers counter that they have limited creative control and can only put an offer in front of the right audience. Why should their compensation be dependent on an advertisers' ability to generate compelling creative, competitively price their products, keep their servers up or have good products?
This is why a complete transition away from traditional CPM pricing, despite many predictions, is not inevitable and is not in the best interest of the involved parties.
If publishers are only paid based on performance, they are inherently bearing 100 percent of the risk of marketing a product over which they have no control. Publishers may accept this risk in the short run for a few well-branded products, but they are unlikely to build infrastructure, commit resources and grow around such an arrangement.
As any tried-and-true direct marketer will tell you, the way to minimize risk is to test, retest and test again. Altering the price structure simply exonerates a marketing department of any responsibility over determining what works and what does not. If publishers accept all the risk, they become nothing more than commissioned resellers.
There is a good deal of buzz about how well online advertising brands a product. Take this simple test: Does "Punch the monkey and win" mean anything to you? What is Bonzi? Do you know "How low do you want your APR"?
CPA deals don't compensate publishers for the huge amount of branding and exposure they generate for advertisers. Many people need to view an ad many times before they actually decide to buy. The value of branding may be uncertain, but it certainly is more than zero.
What is a customer worth?
Advertisers have intimate knowledge of what a customer is worth and how profitable different customers are. The publisher has almost no insight into every advertisers' market segment, and thus no ability to extract a fair market price for a customer.
When the advertiser sets the CPA price, the publisher almost always gets a bad deal. Even with CPM pricing, the advertiser has all the pieces of the puzzle: exactly what the buy cost, how many people saw the ad, how many customers were acquired and how much those customers are worth. With CPM pricing the publisher can objectively compare different advertisers and be appropriately compensated.
Both advertisers and publishers must trust each other regardless of the advertising pricing structure. Unlike impressions and e-mail that can be tracked by third-party auditing services, publishers have no way to audit and validate the sales reported to them. What's more, advertisers are unlikely to give publishers access to the sensitive data necessary to evaluate the risk that publishers bear—even if they had a way to do it.
Big and small publishers
There are two main classes of Web publishers: the big guys with lots of traffic and brands of their own, and smaller players that command less respect. Smaller sites may not have the critical mass or ability to sell their own ad inventory, and must either join an ad network or take whatever they can get. Perhaps performance deals make sense for advertisers here where the cost of buying on so many small sites is prohibitively expensive, but large, branded publishers are a different story.
Some middle ground?
The direct marketing industry has been able to track cells and response rates with tremendous accuracy for decades. So why hasn't postal mail evolved into a performance-based model?
The reason is it doesn't work for the publishers, who own all of the data. The industry has reached a stable pricing level that incorporates risk, branding, customer worth, honesty and size all embedded in the CPM price. The online industry will follow suit with more accurate pricing.
Perhaps current CPMs are too high for some advertisers, but CPAs are certainly too low. To bring advertisers to the table, publishers must continue to provide high quality ad vehicles and uphold online vehicles to the same integrity standards as off-line.
Publishers also have to realize that if advertisers do not sell product they cannot buy advertising. Advertisers, for their part, must accept a fair pricing model that compensates publishers for the full value of the advertising inventory.
While they may be able to cut a limited number of pure performance deals, those advertisers who want to achieve substantial volume and remarket penetration for their products will have to buy advertising with a CPM component.
Seth Lieberman is co-founder and chief executive officer of Focalex Inc., an e-mail marketing company. He can be reached via e-mail at email@example.com.
Share your thoughts on CPM vs. CPA.