Price Discrimination and PPC Ads
PPC is more complicated than it was a few years ago because the ad platforms – Google AdWords, Yahoo Search Marketing, MSN AdCenter, and others – have become more adept at price discrimination. They are better at charging their customers - the advertisers - what the customers are willing to pay. Customers often resent price discrimination but businesses employ it to increase their revenues. We can only expect this trend to continue.
For any product or service, different people have different prices they are willing to pay. If you ever took an Economics course you surely remember the downward sloping demand curve, which is a graphical way of saying that you’ll get more buyers at a low price and fewer buyers at a high price. For a business that cannot price discriminate, this poses a problem. What price to offer? There might be some consumers willing to pay 80, but twice as many consumers willing to pay 50. If you set the price at 50, you get more revenue, but the people who are willing to pay 80 are happy that your offering was 30 less than they were willing to pay. (Economists call this consumer surplus.) The ideal situation for the business would be to sell to some consumers at 80 and others (the price sensitive ones) at 50. Price discrimination – charging each consumer close to what he or she is willing to pay – increases revenue for the business.
Business strategists are forever trying to figure out ways to price discriminate. For commodities it can be difficult, but some markets are conducive to price discrimination. The classic example is the airline industry. Travelers have different itineraries and routes, and the airlines purposely impose complex pricing rules (e.g. cheaper if you stay over a Saturday) in order to price discriminate. Business travelers typically end up paying more than leisure travelers, and if you fly into or out of a small city you pay more than between large cities. On a flight with 100 passengers, it is possible that everyone paid a different price for the seat – 100 different prices for the same product. Consumers often resent these schemes, but economists love them.
Movie theaters price discriminate by charging lower admission for kids and seniors. Everyone gets the same product – a seat in the theater – but consumers that are more price sensitive pay less. Car dealers discriminate based on how much the customer haggles. Sellers of new products, especially consumer electronics, often price discriminate over time. When the iPhone was first released, consumers willing to pay $600 got to buy it. A couple months later, Apple lowered the price and a larger segment of the public was willing to buy. Apple could have charged $400 from the beginning, but then they would have lost all that revenue from the people willing to pay $600.
Buyers often feel like they are being played for chumps when they learn
about price discrimination, but many economists absolutely are crazy about
it and wish we had more price discrimination. Businesses are encouraged
to make prices secret – create a fog of uncertainty – to get customers to
accept prices offered to them. Preston McAfee, an economics professor at
the California Institute of Technology, gave a talk about prices. He raves
about Dell selling the same computer at different prices based on how the
consumer identifies themselves at the website (small business, large business,
home users). He talks about the glory of the airline industry’s blatant
price discrimination. Earlier this year Yahoo hired Preston McAfee to be
part of their marketplace design group. Likewise Hal Varian, University
of California Economics professor who wrote the influential web economy
book Information Rules, is on leave from his academic post and working at
Google, where he has the title Chief Economist. It’s a safe bet these ad
platforms hire economists to figure out how to extract more money from their
Pay-per-click ads sell at auction, but not at the same type of auction used to sell paintings or real estate. In those auctions, the top bidder gets to purchase, but nobody else gets to purchase. And the top bidder only has to pay as much as the second bidder is willing to bid – even if he or she values the property much higher. There is no price discrimination.
PPC systems, by contrast, typically have multiple ads for the same keyword. Not only the high bidder gets to display an advertisement, but also the second and third and fourth highest bidders. While the top position is indeed the most valuable and gets the most user clicks, each click on an ad at another position is just about as good, even though the buyer is paying less for it. By displaying multiple ads that cost buyers different amounts, AdWords, YSM, MSN AdCenter, etc. are able to price discriminate.
This was true to some extent from the beginning – think GoTo in 2001 or so – but over the years the ad platforms become more sophisticated in their discrimination. The early GoTo system had each advertiser pay what he or she bid. The amount bid set the order of ad display, but each advertiser had no direct effect on what other advertisers paid. Overture revised their system to reduce gaming by advertisers; now each advertiser paid one penny more per click than the next lower bid. While this system allowed price discrimination, advertisers were still able to get a consumer surplus. For instance, if the bids were
A - $1.40
B - $1.10
C - $0.95
D - $0.85
The top three bidders would pay (for one click on each ad): $1.11+$0.96+$0.86 = $2.88, while their actual valuation of the clicks was $1.40+$1.10+$0.95 = $3.45. The three advertisers were walking away with $0.57 in consumer surplus because Overture wasn’t able to fully price discriminate. Further, clickthrough rates were not a factor, so a poorly written ad at a high bid could reduce Overture’s revenues by displacing a better written, but lower-priced ad. Further, everyone knew every bid.
Yahoo Search Marketing, as Overture is now called, introduced a new system - Panama – in early 2007. It made the relationship between bid levels and placement much more murky and introduced factors including clickthrough rates and ad quality to the equation. Effectively, this changed the relationship between YSM and advertisers to more of a “trust us” one. YSM bills you after the fact for the clicks; you don’t know ahead of time how much you will pay. You don’t know what other advertisers are bidding or paying. More fog for more price discrimination.
Google’s AdWords program, meanwhile, went into the fog-generating business
earlier than YSM. While at one time their system was a simpler “generalized
second price” auction, they started employing clickthrough rates, ad quality,
and landing page relevance into their ad position ranking and cost per click.
Google’s system further complicates things with cost-per-action, cost per
impression, click-to-call, and positional preference options. Advertisers
have no idea what the competitive landscape is. It’s more complicated than
I went to business school, and do you know what it’s like in business school? You have guest speakers come in and arrogantly criticize businesses for not conducting price discrimination. These loud consultants will scream at the young MBA students: “Yahoo’s advertisers are willing to pay $8 a click and Yahoo is charging them only $6 a click! Yahoo is leaving money on the table!” The MBA students nod their heads and think: Yeah, they’re so naïve at Yahoo; I should tell them how it’s done. Then the MBA students graduate and get jobs at Google and Yahoo and Microsoft and try to impress their new bosses by presenting a case for further price discrimination: “The advertisers are willing to pay $8 a click and we’re only charging them $6! We’re leaving money on the table!” And the bosses nod and say: Yeah, OK, change the system.
And so we get modern, inscrutable ad systems. Advertisers don’t know what other advertisers are bidding or what they are paying per click. They don’t know the formulas used to determine ad ranking, and they don’t know exactly how writing new ad copy or changing their bids will affect anything. They are kept in a fog for a reason: to promote price discrimination. From the perspective of AdWords, YSM, and MSN AdCenter, that’s the way it should be. Each advertiser should pay what he or she values each click at, irrespective of how other advertisers value the click.
The platforms also price discriminate on the basis of advertiser knowledge
and savvy. Contrary to announcements about how easy it is to advertise,
their systems are more complicated than ever. Lazy or unmotivated advertisers
accept the default settings on new accounts, and end up paying much more
than a sophisticated experienced user would. We’ve all seen situations where
a consultant can earn his fee just by telling his client to disable the
content network or change the match options.
Impeding price discrimination
What can we advertisers do to stop this? How can we thwart price discrimination? For starters, we can reduce the amount of information the ad systems get. Don’t use their tracking systems. Both YSM and Google offer website tracking systems. You have to be a YSM advertiser to use their system while anyone can sign up for Google Analytics. Google further offers “Google Checkout,” a shopping cart and payment system for the small business. These systems allow the platforms to collect information, which can be used to estimate the value of a click to the advertiser. Yahoo says they use the information to determine which affiliates are good and which should be subject to Quality Pricing discounts. Maybe that’s true, but if the ad platforms have conversion data, it opens up the possibility for them to adjust prices so each advertiser pays what a click is worth to him or her.
This is not a paranoid fantasy; it’s a real possibility. And even if the ad platforms don’t estimate values at the micro level for each advertiser, you can bet they are doing it at a macro level for all advertisers in an industry. They no longer rely on simple auctions, which were insufficient to extract the full click value from the advertisers.
In my view we all have a duty, as citizens of the web, to degrade the ad
platforms’ ability to estimate click values. Don’t use their tracking systems.
Don’t let them have conversion data. Don’t just do this for yourself; do
it for everyone. Don’t let Google or Microsoft or Yahoo or anyone else know
how much those clicks are worth. It’s none of their business.