home > websmarts
Advanced Online Advertising

Other articles by Seth 

By Seth A. Maislin

Originally published at Webmonkey.com.
Considered for republication at Adverising Express (India).

  I'm a content guy with stock options. My job is to make things findable without search engines. I put sensible links in sensible places. For this I get paid the big bucks.

  People ask me where this money comes from. "And don't say ads," they command. "I never click on ads. Nobody clicks on ads. We hate ads. Ads are useless and obnoxious. You can't be making money from ads. Right?"

  Methinks they protest too much. Advertising is still the biggest and fastest way to make a company money. The Internet is still a media industry. The Net was built by capitalists. If there are other ways to make money, people are doing it: subscriptions, reach partnerships, content partnerships, online commerce.

  Of course, to make the big bucks (like me), you still need sensible links in sensible places. If you want to get rich, you need to be intelligent about link placement.

Making an Ad Impression

  Okay, we've seen the friggin' ads. Loosely defined, online advertisements are marketing-oriented hyperlinks. They don't require an image. Advertisers pay site owners for the space to display the ads; we call this space "real estate," and the ads "creatives." Advertisers produce the creatives, and site owners price the real estate. For many sites, advertising revenue is the sole generator of income, as it is for the entire media industry.

  There are three pricing models for online ads: impressions, ad clicks, and conversions.

  Let's start by talking about impressions. Impressions pricing is a pay-per-view model. Every time an ad is displayed on a page -- "impressed" upon the page -- the page owner can collect money. In this model, it doesn't matter if a single person sees the ad 1,000 times or if 1,000 people see the ad once each: It's still 1,000 impressions. It doesn't even matter if they notice it, let alone click on it. If it appears on the screen, it counts.

  Here's a laugh: Impressions don't impress. Usability studies demonstrate that users do stop looking at ads. Some users go out of their way to hide them by scrolling them off the screen, installing ad-blocking software, and covering the monitor with their hands. Noisy ads are too noticeable and annoy people; subtle ads are ignored without prejudice. So what's the point, right? These days we'd expect impressions to be pretty worthless -- and getting worse.

  A mindless, run-of-the-mill impression is worth one-tenth of a cent, at most. As the Web grows, the supply of real estate skyrockets while the demand for it plummets. There aren't enough creatives in the world these days (frightening thought!), so many websites display self-promotional "house ads" just to fill space -- vagrants living in new homes.

  For now, ad impressions are still the largest source of income for highly trafficked websites. If a website can generate 10 million impressions per day, it can make over $3,500,000 per year. The Terra Lycos network (Webmonkey's parent) generates billions of multi-ad-toting pages every month.

  Odds are that your site doesn't have 50,000,000 visitors per month like Yahoo. So if you want to make millions, you might want to increase your audience, or "reach." But growing reach is extremely expensive and requires campaigning and acquisitions. Fortunately, there's an alternative: targeting.

Selecting Your Target

  Which is more valuable, a billboard on a major highway or a street sign on an obscure access road?

  To express your idea to millions of people, rent the billboard. Then again, who cares about a billboard at 6 a.m.? Commuters might ignore it as part of the scenery. A billboard on an access road is even worse, because only 100 people are there to ignore it.

  Consider an access road sign that says, "Last Fuel Station for 50 Miles." I'd bet lots of people pay attention. I'd also bet a lot of people buy fuel because of it, too. This sign is well targeted, in that it communicates an immediately relevant message to a precisely appropriate audience. Targeting is the process of smartly connecting products and services with the most desirable and likely consumers.

  Web-based targeting can be static or dynamic. Static targeting means choosing real estate based on its predetermined audience demographic, such as selecting a personal fitness website for ads on exercise equipment. Dynamic targeting means choosing real estate via real-time inspection; this requires more interesting technology. Geographic dynamic targeting, for example, occurs when the geographical location of the user is considered before Web pages are displayed. By reviewing an IP address or stored postal code, specialized ads can go to users simply because they're in Denmark or at Harvard University.

  Dynamic targeting can be implemented as a result of site personalization or customization, such as through user registration. Advertisements can be displayed to users who have bought similar products, work part-time, live in California, or read Wired. Any website that can publish "Welcome, Seth!" on its homepage can serve me specialized ads. Ads can be specially suppressed, too, including ads for products already purchased or with content inappropriate for kids.

  If you do it right, if you truly understand how targeting works, you can replace that run-of-the-mill ad with a brilliantly appropriate ad worth 10, 100, or 1,000 times more per impression. If you can build a system that identifies Spanish-speaking mystery-lovers living in Florida, you can charge more than $1 for a single impression. Just don't forget how much smaller a targeted audience can be.

Stop Staring and Click

  Two pages ago I said there were three pricing models, and one of them is based on impressions. A second model measures clickthrough or ad clicks, which is the number of times an advertisement is clicked on. Some advertisers don't care about clickthrough and instead remain interested purely in being seen and becoming household names. But many Web-based advertisers want to draw in customers and increase reach. Clickthrough is more valuable than unclicked impressions -- 10 or 100 times more -- so site owners charge more. Many advertisers pay fees for both impression counts and clickthrough percentages. Also, clickthrough goals provide a financial incentive for site owners to give these ads excellent visibility.

  My banner-hating friends claim to never click on ads, and it's true. Clickthrough percentages for untargeted ads are small, often less than 1 percent, and still dropping. More than 100 impressions are necessary before a single user responds to an ad. Still, whereas impressions are worth $0.001 each, ad clicks can be worth $0.10. With targeting, ad clicks imaginably can be worth $10 each!

  Site owners have a clear financial incentive to improve clickthrough. In fact, clickthrough rates are a good measurement for promotional efficacy. Quality products with catchy advertisements earn the highest number of ad clicks. Then again, so do ads that trick users into clicking on them, like those hateful input boxes that are really images.

  Just because somebody clicks on an ad doesn't mean that user spends money (or time) at the advertised site -- especially when users are deceived into clicking with fake user interface elements, or when additional browser windows are pushed without warning onto the user's screen. Awarding $10 to a referrer just because they fooled an unsuspecting non-customer into clicking the ad is pretty stupid. So to protect themselves, advertisers have a third pricing model: conversions.

  Conversions are completed sales or transactions that occur as a direct consequence of advertisement-based referrals. A user who clicks an ad and then buys something or signs up for a mailing list has been "converted" from a visitor to a customer. The conversion model supersedes both the impression and ad-click models in that impressions and ad clicks are virtually ignored. Money changes hands only when conversions happen.

  Referral-conversion relationships are a type of affiliate program, which might pay referrers a percentage of the complete sale value. (The site owner is considered the "affiliate.") For example, if my website caused a user to spend $400 at a friend's website, my friend might reward me with 5 percent of the purchase, or $20. Giving away a percentage of gross sales is expensive, but it's often cheaper than advertising. Further, conversion-based ads often are accompanied with positive, word-of-mouth reinforcement. Since affiliates earn money only when a sale is completed, driving a user to complete a transaction is the important goal. In this way, the responsibility for good salesmanship -- and good targeting -- rests on the affiliate.

  If you want to know more about getting commissions on referrals, check out Webmonkey's Affiliate Network Lowdown article.

  There's one funny wrinkle in the conversion model. No matter how well the ad is written and targeted, if the product or even the transaction experience itself is terrible, no conversion will take place. This is like convincing someone to eat rotten food or patronize an out-of-business restaurant.

  If you're an advertiser hoping to find affiliates, make sure referrals actually lead to completed transactions. There are too many companies out there with brilliant ads and horrible websites. A great ad with a horrible user experience is twice a waste of money.

Subscribing to Content

  My curmudgeonly friends, who hate ads more than Florida mosquitoes, say they're willing to pay money to make ads go away. This is an example of a subscription, where money is paid for access to exclusive or special content or services. In my friends' cases, "high quality" means "no ads." Subscriptions are a great way to ensure that Internet money keeps flowing, because most subscriptions renew automatically. Unwillingness or failure to pay subscription fees means missing out on service and content possibilities.

  Subscription fees depend on the quality of what's provided: High quality begets high rates. Subscription payment plans can be periodic, pay-as-you-go, or part of an exchange agreement. Periodic fees are the most common, where the user pays some amount after a regular time period. My ISP charges me $20 monthly, for example.

  Pay-as-you-go subscriptions are much like e-commerce itself. Users pay a small amount of money for continued access, like quarters into an arcade game. Effective models include charging $15 for a stock brokerage transaction, $1.50 for a file download, 5 cents for a financial transaction, or the mythical micropayment model. Users might be prompted to prepurchase a number of transactions, depositing money into an account and then spending that money a little at a time. Pay-as-you-go subscriptions always have sign-up requirements, which is why they're different from other e-commerce.

  Exchange agreements are strategic deals in which no money changes hands. For example, a newsletter subscription might be awarded free to a contributing author. Services are traded for services, advertisements for advertisements, content for content, etc. Since I'm writing this article about Web money, and since exchange agreements don't have any, you might think I'm going to skip over them. Not true. You'll see why in the next section.

  But wait, there's more. Subscriptions mean mailing lists, and mailing lists are worth money. All the information my ISP has is saleable, depending on their privacy policy. Often personal information is collected in exchange for free services, because that information is later sold at a profit -- and users on the list get junk mail.

  And that's pretty much it as far as making money on the Internet goes: You sell ads, or you sell a subscription. Here's what's funny about this. There are a lot of users out there who cringe when they think about ads. There are also a lot of users who demand that the Web be totally free. If you're hoping to earn money, you have no choice but to choose, 'cause you sure can't do both.

Partnerships and Reach

  Let's say you decide to build an advertising network or you devise a sensible subscription plan. The battle is only half over. The other half is audience. Without audience, ads and subscriptions are useless. What's worse, audience is very expensive.

  How expensive? Remember when Excite bought Blue Mountain Arts in October 1999? It cost them $750 million. BMA offered a free service, required no registration, and served no advertisements. Translation: They made no money. But Excite decided they were worth close to a billion dollars! Why? Because one million unique people visited the website every day, and that's a lot of people. (The industry refers to viewers as "eyeballs," so increasing one's reach is about "finding eyeballs.") Doing a little math, we see that in 1999, one unique daily visitor was worth $750. If you were to visit the same website every day, maybe to check e-mail, you would have been worth $750. At those prices, if you visit Webmonkey once a week, you're worth $109. Cool for Webmonkey.

  It's misleading to use a long-ago purchase as a metric, but anecdotally it illustrates the value difference between people and ads. On that takeover date in 1999, a single daily visitor was equivalent to 750,000 untargeted ad impressions. As the supply of Web users increases, this number drops.

  If people are worth so much more than ad impressions, it doesn't make sense to buy ads. It makes more sense to buy people ... I mean, eyeballs ... I mean, reach.

  There aren't many big players in the industry anymore. It's expensive to buy access to millions of unique users. Instead, companies elect to "borrow" the eyeballs from companies that have them. Put another way, companies can subscribe to an audience. These are partnerships in which a company pays for access to another company's audience. A company with great content and no audience agrees to pay a company with a huge audience but no content. Companies with audience that search for content partners are known as "content aggregators." What's bizarre about the Web world is determining which is more important: content or audience. If 100 visitors are worth $1,000, how much content is needed for a fair trade?

  Without good content, there are no visitors. Without visitors, content is effectively valueless.

  The challenge for most websites is to accept some balance between the two: content that is good enough to keep users, but not so expensive that the website loses money.

  Money exchanged between partnered companies can move in either direction. Some content aggregators buy high-quality content, hoping to draw in a larger audience; some content producers buy space at high-trafficked sites, hoping to find consumers. Regardless, the spending company is looking for eyeballs because without eyeballs -- without subscriptions, ad-clicks, and impressions -- there's no money.

  To understand this, let's look outside the Internet world for examples. The pay-for-content (or pay-for-service) model is clearer: Bakers get paid for bread, mechanics get paid for fixing things, and inventors receive royalties for their inventions. The pay-for-audience model is not as clear because it looks a lot like ordinary advertising: Soft drink companies supply clothing for sports celebrities, publishers supply end-of-aisle displays in bookstores, and AOL gives free CD-ROMs to people who buy computers.

  It doesn't matter if audience is bought, or if content is bought in the hopes of earning audience later. Partnerships are about eyeballs. Once a company has huge reach, it can command more money for advertisements.

The Whites of Their Eyeballs

  So there's this startup with a great product. They are bought by a company with a huge audience. Is this a marriage made in e-heaven? Suppose the parent company is a law firm with hundreds of clients, and the startup makes yogurt.

  Ads that reach the wrong viewers are as bad as those we cover with our hands. Eyeballs don't matter if they're the wrong eyeballs. On the other hand, as etown.com CEO Robert Heiblim once said, "Getting the right message to the right audience always works regardless of the medium."

  Maybe this sounds obvious, but it's easy to be overwhelmed by the Internet's size. Remember the Blue Mountain Arts purchase? I estimated that a single daily viewer was worth 750,000 untargeted ad impressions. A few pages earlier, I argued that targeted ads can be worth 1,000 times more. In other words, a BMA daily visitor is worth 75 brilliantly targeted ads. But just how brilliantly can Excite target one million unique viewers each day?

  Imagine a company called Uphill Shipping, which provides custom luxury shipping solutions to a select group of 100 clients. One client is an art gallery in New York City that occasionally sends irreplaceable artwork to a sister gallery in Paris. Artwork like this has to be specially insured -- a perfect opportunity for Dundroppit Insurance. If Dundroppit wants to increase its reach within this select market, they should consider a partnership with Uphill Shipping. If Dundroppit becomes one of Uphill's listed insurance providers, it could gain as many as 100 new clients, and each client is a guaranteed source of high income. In fact, Uphill's loyal clients are probably more valuable to Dundroppit Insurance than millions of unknown Blue Mountain users were to Excite.

  Again, this is targeting. Targeted partnerships are way more valuable than the untargeted kind. The more you know about an audience, the more value it has. The demographics of Uphill Shipping are extremely specific: not-for-profit institutions for lovers of artistic masterworks, funded by public grants and private donations. This is exactly what Dundroppit was looking for.

  Exclusivity arrangements are icing on the cake. Certainly companies can partner with competing companies (as long as the companies don't mind). Bookstores, for example, are implicit partners with hundreds of book publishers. Nevertheless, exclusivity is a big part of the growing Internet as competing companies struggle to find and keep new users.

  The best partnerships are both targeted and exclusive. The more a company knows about its reach demographics, the more valuable those eyeballs are, and the more money it can command in advertising deals, subscription plans, and other partnership arrangements.

Show Me the Money

  Let's follow the money. Company A pays for a targeted audience by buying a mailing list, and then Company B buys targeted advertising space from Company A. Company B got its money by selling advertising space at its own site to Company C. C sells space to D, D to E, and E to F, and sooner or later one of these companies will make money by serving ads from the first company, Company A. Sure, partnerships pass funds between companies, and exclusivity keeps money away from competitors. But there's no new money.

  Subscriptions bring in money, but this money usually gets spent on overhead like content and infrastructure. In an industry where users are less and less likely to spend on content, subscriptions aren't enough, online periodicals are dying, and it's harder to find partners who pay well for eyeballs.

  The largest source of new funds is from venture capitalists (VCs), and that money is spent on infrastructure, content and marketing. The investors claim the new audiences, then sell their eyeballs to other companies. Basically, VCs give money to the VCs before them, and for each market niche, the last VCs to join the fray lose.

  After company-to-company transactions (ads and partnerships), subscription money, and VC-to-VC money, there's only one money source left: e-commerce. Simply put, e-commerce is buying something over the Internet. It's not special or spectacular. It's an ordinary financial transaction, governed by the same rules as the rest of the world, only without a cash register. Good products and services at reasonable prices are bought by consumers who know about them.

  E-commerce is comprised of three kinds of sellers. First come the people who make and market the products and services themselves, for a profit. Next are the middlemen, who build the e-commerce software or enable the transactions to take place. Just as superstores and supermarkets provide access to multiple products from numerous companies, so do online auction sites and B2B companies enable all the online buying and selling. The last sellers are the affiliates, who receive referral kickbacks when a transaction is competed. In all cases, money enters the system from the end-buyer and gets distributed to everyone involved: some to the referrer, some to the middlemen, and the rest for overhead and profit.

  I don't think there is any new money in the Internet. All this hype about IPOs and B2Bs is just an illusion. The virtual world is simply the real world on telephone lines. We communicate and negotiate as before, buy and sell like always, and pass dollar bills to each other.

The Big Winners: Us

  Your mission, if you choose to accept it, is to have more money in your pocket than you did yesterday. You have two tools at your disposal.

  The first is targeting. If you can target your advertisements, you can charge more. If you can target your subscriptions, you'll get more subscribers. If you partner with the right companies, you can gain exclusive control over the most promising customers.

  The second is infrastructure. The unsung heroes of the Internet world are the companies who build the tools on which the Internet runs. Certainly it's less glamorous. Nobody likes to talk much about routers. Try convincing someone that waste management or water purification are sexy, and you'll see what I mean. But it's got to be done, and as long as money is going to change hands the infrastructuralists are going to get a healthy piece of it.

  Here's where we come in, the human beings. We do both, the infrastructure of targeting. The fancy phrases are "information architecture," "interface design," "integration management" and "usability engineering."

  We put links where they're supposed to go, and hopefully that puts the green in our wallets.

Other articles by Seth 


Copyright 2000, 2002 Seth A. Maislin


Site design by little graphics studio.
© 2002   All rights reserved.