JP Morgan Sees Long-Term Dominance For Performance-based Ads; Online Video Loses Luster

By David Kaplan - Mon 05 Jan 2009 12:52 PM PST

It’s not surprising that during a downturn, the clear metrics and ROI offered by performance based ads are looking more attractive. But in his wide-ranging ‘09 outlook, JP Morgan analyst Imran Khan expects marketers to treasure performance-based ads even when the larger economy begins to grow again. So the market share gains that performance ads have achieved over the CPM-based model look pretty durable and mean continued struggles for display ads.

The report, Nothing But Net: Outlook for Global Internet Stocks in 2009 (PDF), predicts that the mostly performance-based U.S. search ad market will rise 10 percent in 2009 to nearly $16 billion. In contrast, display ads, which includes both performance and branded advertising, will grow only 6.3 percent to $8.4 billion this year.

—Bearish online video: Khan expects the accelerated shift to performance ads having a dampening effect on the growth of online video ads. Even in a series of downward revisions, online video ad growth still seemed poised for healthy gains this year. For example, at the end of November, eMarketer forecast online video growth of 44.9 percent, which was still nearly half of the 81 percent growth rate the researcher predicted for 2008. Khan believes that online video is headed for a considerable slowdown because one, it’s still reliant on the CPM model, as opposed to performance-based measurements like cost-per-click or cost-per-action based display. And unlike television, which still can count on advertisers to respond to CPMs, online video can’t guarantee viewership for any specific video the way TV does in the upfront model. Plus, considering the unpredictability of popular videos, the uneven quality, and the continued battles over copyright, JP Morgan doesn’t expect online video to have great prospects for the next few years. However, Khan is intrigued by Google’s experiments with an e-commerce platform—i.e, performance-based model—for YouTube videos. For example, if a user watches a song featured in a music video, they can click on a link that lets them buy music directly Amazon (NSDQ: AMZN) and iTunes, with YouTube getting a cut of the revenue. More after the jump

—Social nets need new approach: Looking at the projected slump in online ad sales for sites like Facebook and MySpace, JP Morgan strongly advises that online communities look for sources other than display for revenue. In particular, the report offers a few possible routes to profitability, including greater use of cost-per-action ads, lead gen, sales of virtual items, connecting to classifieds or e-commerce sites, and charging for premium membership, as LinkedIn and Classmates do.

—Mobile looks good long-term: But as for the near-term, the long-awaited explosion for mobile ads will simply have to wait for a better economic environment—not to mention better phones and technology, Khan says. although mobile phone penetration is high at 84 percent in the U.S., the mobile search market is in the early adoption stage. In Q108, only 15.6 percent of wireless subs were using mobile web, according to Nielsen Mobile data. Even within this small subset of mobile internet users, usage drastically trails that on PCs. Nielsen Online says that a PC online user visits more than 100 domains per month, whereas mobile web users visit 6.4 individual sites per month, on average.

—M&As start slow, but gather steam in H209: With last year’s total deal count down by 20 percent, according to a report last week from DeSilva + Phillips (disclosure: one of our sponsors), JP Morgan anticipates continued coolness in the acquisitions area for at least the first six months of 2009. But things are likely to heat up in the second half of the year, assuming the economy begins to stabilize.

Google Launches Affiliate Advertising Network, Courtesy of DoubleClick

Erick Schonfeld

36 comments »

Amazon, watch out. Earlier today, Google launched an affiliate ad network. Or, rather, it rebranded Performics, the affiliate ad network that came along with its purchase of DoubleClick, as the “Google Affiliate Network.” As with other affiliate networks such as Amazon’s, participating Website publishers get paid a fee for each referral that results in a sale. Existing advertisers include Bank of America, Barnes & Noble, Citi, Target, and Verizon.

The service isn’t yet integrated into Google AdSense (publishers and advertisers still have to set up separate accounts), but that would be a logical next step. An integration with AdSense could add a contextual element to the affiliate ads placed through the network. The more relevant Google can make those affiliate links, the more that consumers will actually click through and buy (in theory).

Google also continues to experiment with a pay-per-action advertising program, which is still in beta. At some point, it might make sense to consolidate that effort into the Google Affiliate Network as well.

Update: Google will actually be phasing out the PPA program at the end of August as part of the integration with DoubleClick. You can read more details at the blog post here.

Horton hears a Microhoo

If the acquisition goes through, it will alter the primal core of online advertising and elevate the value of CPA. Here's one VP's predictions.

It's the biggest business story in months: the Birth of Microhoo (or Newshoo). Online advertisers and those who serve them have been spellbound by every one of the 63,387 (and counting) Google News entries about Microsoft's (so far unsuccessful) attempts to get a little Yahoo in its system. Why? Simple. Microhoo will alter the primal core of online advertising. In fact, it has already.

Let's say it together. Online advertising is about results. Before we all merged onto the Information Superhighway we tried to grab as much SOV/SOM as possible and hoped the message would motivate audiences to do…something. Now, with the net's instant metrics, advertisers track the success and failure of campaigns as they happen. This gives us enormous quantities of user data, and with it, advertisers deliver relevant ads that convert to sales. Which leads us to the importance of Microsoft's attempt to buy Yahoo.

Yahoo knows (almost) everything about each one of us, and so Microsoft's play is to grab that information and use Yahoo's platform to cripple Google's stranglehold on online advertising. Without Yahoo, Microsoft doesn't have all the pieces to chip away at Google; Yahoo's platform technology and user data feeds Microsoft's business goals.
Microsoft used to be the one big fish in a pretty small pond. Unfortunately for it, gigamouth (Jasconius pelaganax) Google jumped in and Gates' behemoth no longer rules the seas. Yahoo, on the other hand, was surely the first internet darling, but is now coping with a dam (bubble) bursting and big G's search and online advertising dominance. This is the meat and potatoes of the proposed merger, as Microsoft needs to acquire a Yahoo (or, gasp, AOL) to out-swim Google.

The strangest of all is Google's offer to help Yahoo. As AdWeek succinctly states, it's "putting Yahoo in the unenviable position of having to embrace the help of one contentious rival in order to fend off the advances of another."

With AOL on its way out, consolidation is sensible for Yahoo -- especially with Microsoft's ever-deep pockets and drive to remain relevant. We've seen a ton of big pairings in the past 18 to 24 months (Google/Doubleclick, AOL/Tacoda, Yahoo/BlueLithium) and with these happening, those companies providing distinct ad models are trying to demonstrate that theirs is the future.

So which online advertising will Microhoo use? It's a particularly fascinating question in the context of its mammoth share and reach. According to comScore, a Yahoo and Microsoft combo would have 32 percent of the U.S. search market, while Google holds 59 percent.

Yahoo has about 75 percent reach (pretty astounding); however, when it comes to user engagement, the company's numbers hover around 15 percent. Microsoft can take Yahoo's reach and employ a guaranteed means for advertisers to convert leads into customers: cost-per-acquisition (CPA).

CPA is not new, but all these consolidations mean CPA will be the darling of online advertising. Microhoo and Google's advertising customers will demand guaranteed results and they will be compelled to provide them because there will only be two fish swimming around -- and the choices of the smaller ad models will diminish. Since CPA advertisers only pay when based on predefined customer actions, Microhoo -- using CPA -- immediately has a leg up on Google's soon-to-be-archaic model, which is susceptible to click fraud and suffers from lack of transparency.

Speaking of click fraud, this shotgun marriage will affect pay-per-click (PPC) rates as well. Advertisers who use PPC are charged every time a user clicks on an ad; Microhoo will try to leverage that business. If these rates rise, advertisers are going to move swiftly to a model -- like CPA -- that has lower rates and higher returns.

CPA will in fact deliver high conversion rates for Microhoo's advertisers, and CPA's scalability across a large network means Microhoo can focus on integrating different ad exchanges and search ad systems into a more robust and powerful network. Remember that Microsoft is still integrating aQuantive -- which brought Atlas ad-serving technology, Avenue A/Razorfish agency and DrivePM ad network -- into its system. Incorporating a large CPA network will dominate rather than buttress ad models like behavioral targeting, contextual and search that are falling by the wayside.

When all else fails, simply follow the money. Microsoft employing Yahoo's platform with a CPA model will bring in the major bucks for shareholders, for the behemoth, for advertisers. Google has tried to deploy CPA, but because it makes money in search, CPA has not gained traction with Google's clients. So the question is: Will Microhoo's eventual use of CPA mark the beginning of the end for Google? And who will end up in the deep end?

CPA and The path to social network riches

By Mason Wiley

More by this Author
Contact Author

Advertising on social networks can be very expensive -- with little payoff. Here's an easy solution to this cost problem.

Social networks are commanding attention -- for good reason. Millions and millions of people visit them daily (and to many an employer's chagrin, hourly). Given the sheer number of users and their length of stay, the big social networking sites like MySpace, Facebook and others should be an advertiser's paradise. But they're not -- yet.

Ironically, the huge number of impressions generated by social networks poses challenges for advertisers. For example, buying all those eyeballs on a cost-per-thousand (CPM) ad model can be prohibitively expensive for many. And unlike sites centered around a specific interest, social networks attract so many different types of people that targeting is difficult. Behavioral targeting offers promise, but hasn't yet evolved to the point at which it can guarantee a positive ROI.

A new ad model has emerged that can. The cost-per-acquisition model (CPA) provides advertisers with a guaranteed way to ensure ad efficiency, while also opening the door for social networks to monetize more of their traffic.

The benefit of CPA for advertisers
The CPA model is ideal for social networking sites because it eliminates virtually all the risk for advertisers of buying on a CPM basis. As a performance-based model, advertisers pay a fee only for the results their CPA campaign generates. That result can be any transaction specified by the advertiser. For example, an advertiser places a CPA campaign on ESPN.com. Rather than paying for all the people who see that ad, the advertiser only pays the publisher for each user that not only clicks on the ad but also follows through and completes the desired action as defined by the advertiser -- anything from an email submit or qualified lead to a sale or paid membership.

Through its 1:1 ratio of pay-per-action, CPA changes the rules for advertisers. The need for narrow targeting to eliminate ad waste becomes irrelevant since there is no waste. Similarly, there's almost no possibility for click fraud since advertisers only pay for results. And unlike CPM, the results of a CPA campaign are directly verifiable through tracking technology, which enables advertisers to determine which ads convert and which do not. For example, an A/B test of landing pages can be measured against the actions completed via any number of creatives or landing pages as well as the conversion processes. The path to purchase is easily tracked back to the source, including the placement of the original creative.

But the single greatest appeal of CPA for advertisers is that it is performance-based. With advertising costs directly tied to results, it provides them a way to advertise with full accountability and control. Advertisers simply establish the actions -- signups, sales, leads, etc. -- they want to reward, and establish how much they're willing to pay. If online publishers think the campaign will generate sufficient response with their audience base to meet revenue goals, they can choose to run the campaign.

Certainly, by running campaigns based on the promise of a potential future reward, publishers are taking a risk. But given that CPA campaigns are more aggressively focused on generating response than the many revenue-sharing campaigns available, it's a gamble that has been paying off for them. Now, social network publishers are finding that CPA can pay off for them, too.

The benefit for social networks
Like any publisher site, social networks want to maximize advertising revenue. Cost-per-thousand is the most prevalent model they use to do it, and it seems to make sense because CPM focuses on the number of impressions generated -- and these sites generate a lot of them. The problem is that these impressions come from an incredibly broad cross-section of the population, many of which can be irrelevant to a given advertiser. As mentioned above, although these sites generate billions of impressions, targeting the right ones is challenging. While targeting users based on where they have clicked (behavioral targeting) can help improve the odds, it is by no means an exact science or a guarantee that a sufficient number of users will perform the desired action.

These inherent challenges under the CPM model have limited the appeal of social networks for advertisers. The result has been a tremendous surplus of excess inventory that despite all best efforts just can't be sold. And a lot of those billions of impressions have gone un-monetized. But now that's changing thanks to CPA.

With CPM, the advertiser assumes the risk by paying the publisher for uncertain potential results. In CPA the publisher assumes the risk by running ads for uncertain potential payments. But when its remnant inventory at issue, the publisher runs no risk; it is inventory that would otherwise go unsold. So it only stands to reason that a social networking site will make more money running CPA ads on remnant inventory than it can just selling space based on CPM alone.

Conclusion
The CPA model is not new, but it has evolved to be a highly cost effective way for advertisers to use social network sites, and an effective way for social networks to generate increased revenue. And new widgets and applications are emerging for social networks that can make CPA even more powerful in the future.

Worldwide Expansion of Google Pay-Per-Action (CPA) Beta

Posted: 21 Jun 2007 02:57 AM GMT-06:00

Google announced today via a press release that they expand the Google Pay-Per-Action beta program to include beta testers around the world. As of today are willing advertisers and publishers outside the United States able to participate and apply for a beta account. Quote:

“Starting today, advertisers in the beta will see an alert in their AdWords account informing them that they can now create pay-per-action campaigns. Going forward, advertisers who have enabled AdWords conversion tracking and received more than 500 conversions from their CPC and CPM-based campaigns in the past 30 days will be automatically added to the beta on a rolling basis.”

(more…)

How To Avoid The Pitfalls Of CPA Marketing Online

When advertisers choose to engage in CPA (cost per action or acquisition) advertising online, there are several key areas that require specific focus to protect their brand and to acquire the highest quality of new leads through these channels. Although advertisers are often tempted to jump in head first, they should not do so without careful planning and due diligence to understand the space and players in it. In this model, the advertiser eliminates the upfront media risk and avoids payment based on impressions served or clicks generated, which do not take into account the number of orders actually delivered on a given media buy. However, the downside of this performance-based model is that advertisers take the risk of what quality of customer or prospect they will pay for. The backend retention or churn rate becomes extremely critical, as advertisers evaluate whether this type of acquisition model is effective and scalable for long-term integration into their online marketing mix. An advertiser should approach this model with clear objectives, defining and optimizing performance metrics specific to the channel selected and the economics to make it profitable. This will take into account both the CPA paid at the time of acquisition and the quality of that consumer generated. Below are the most important strategies that should be used by an advertiser who wants to take advantage of the very real and scalable volume available in the CPA marketplace without falling prey to the many risks that are also an unfortunate and inherent characteristic of this model. Know and trust your CPA agency and/or media partner(s). The CPA marketplace is virtually a "blind" environment for advertisers in terms of what sites and types of placements will be used to drive new customers. In exchange for eliminating the media risk, advertisers give up much of the control over the actual media placement side of the equation. However, if you are working with a trusted partner(s), this will become a non-issue and the partner(s) will guide and educate you on how to utilize the channels successfully. Limit the number of broker/network partners. Whether you decide to go the route of an agency partner or work with the media providers directly, it is best to limit the amount of media partners that you engage. This is particularly true of broker and large network partners. Why? There is a tremendous amount of overlap and cannibalization that occurs at the sub-affiliate level where these brokers and networks generate most of the volume for offers on a CPA basis. The logic of "the more people I have working on my behalf, the better" does not hold true in a CPA model. It actually creates a disingenuous situation for marketers and can have detrimental effects on the advertiser's ability to control pricing and volume in these channels. A single or sole source partner model will almost always yield the highest level of benefits (both cost & volume) to the advertiser if the right partner is chosen to manage this environment. Use real-time credit card processing and data validation techniques during customer acquisition/sign up process. One of the biggest ways an advertiser can protect its business against incidences of fraud or payment delinquency by the consumer is to process credit cards in real time, instead of performing an authorization or Mod 10 verification process at the time of sign-up. The highest level of credit card validation with your merchant processor is the best method for confirming a true buying consumer. An example would be checking the credit card billing address against the card holder's name. Many advertisers that fail in this environment do so due to a lack of real-time processing of payments for initial and recurring credit card transactions and/or a lack of data validation (valid address, city, state, etc.). A key to minimizing fraud in this environment is to only accept orders and customers that you can verify without any delay or offline processing. Put parameters on your affiliates. There should always be parameters placed on your media partners in this channel. Do not go into this channel allowing free reign for your affiliate partners, or you will likely get burned. While "incentives" has become a dirty word over the last 18 months due to quality and fraud issues in the marketplace, there are legitimate strategies and incentivized media partners that, when used correctly, can effectively drive new customer acquisition efforts. It is up to advertisers to perform basic research and to put limitations on their affiliates. Two areas I recommend focusing on are the rules of engagement for the use of paid search on branded terms and the use of pure "cash back" incentive sites. These tactics can lead to costs that surpass acquisition costs and higher cancellation rates for new customers. Create an offer that protects you specifically for this channel. Advertisers must create a compelling introductory offer that will drive the highest level of conversions and compete effectively in these channels; however, the offer should also protect the advertiser should the consumer cancel or not convert beyond the initial transaction. Consumer behavior is more impulse-driven in this environment versus other channels, like paid search. An advertiser should create an offer that balances the needed "skin in the game" by the consumer with a reduced first month charge, or has shipping and handling that also covers the upfront processing charges or the cost of goods sold. This protects the company should the consumer cancel after the initial transaction, and would result in, at minimum, a break-even position. When all is said and done, advertisers will never be able to completely eliminate the risks that are inherent in the CPA marketplace, but by following the above guidelines, they can minimize these risks and devise strategies that will create an equitable balance between the risk and reward. Angie McCloskey is Vice President, Business Development of SendTec, Inc.