|
|
 January 1999
|
|
By Robbin Schindele
It doesn't make sense until you remember the commercial Internet credo; "All that matters is traffic." Traffic is the name of the game. People have to want to visit your site. If you can convince them to do so buy giving away a product or service they perceive has value they will come in the tens of millions. That's why search engines like Yahoo! and Infoseek have so much clout. Everyone uses search engines and everyone uses the big ones. In fact, the five biggest search engines do 98% of the searches. They have traffic. But what do they have to sell?
Your eyes. In the instant it takes your eyes to scan a Web page for the information you want your eyes also pass over an ad. They're selling that brief moment in time and perception. Not a surprise and not a new idea, give away the product (entertainment or information) for free to millions. Then charge companies to put their name in front of your viewers. TV has been running that way for decades, and before that radio and newspapers. It makes sense that the first model a hugely popular medium like the Internet/WWW would turn to is one that has worked so well in the past.
The Internet though presents a very unique problem. It's a voluntary, interactive activity. Television and radio are both passive media. The viewer/listener contributes nothing to the experience except their attention. With the Web users must do something. And since they're involved anyway, they don't have to do anything they don't want to, like click on your ad.
There is a unique technological problem too, speed. The Internet/WWW is a different medium for different participants. They all don't have the same experience using it. The slower their connection, the more congested their ISP, or the backbone or the servers they're trying to access, all affect how they perceive the Internet. Anything that can be done to make the process more seamless, more instantaneous is good. Graphics, sound and music, interactivity, all the attractions the World Wide Web has over the text based Internet, take time to download and time to display. A slow computer on a slow connection makes adding anything of any size (big graphics, audio files, an interactive game) tedious. So how do you put advertising on a site when the best way to do it (the best for all viewers) is to make it simple and small?
When the web started to explode in 95/96 it was a real problem. People were complaining about the web being slow already and now the pioneers wanted to start getting some payback on their time and cleverness investment by putting ads on the pages people wanted. Making text based ads didn't appeal to advertisers because they had invested millions and sometimes decades nurturing and cultivating graphic representations of their products, the Coke bottle, the yellow and orange Tide box. That was what differentiated them from their competition. They were the symbols, icons, shapes, or names people knew and trusted. If they wanted advertising dollars, publishers had to find a way to give customers what they wanted. Nobody would know it was an ad if it looked like just another ubiquitous blue link. So Wired magazine invented the banner.
It's kind of hard today to imagine that someone had to invent the Banner ad. For the past three years it has been the workhorse of Internet/WWW advertising. Small, the standard is now about 60 pixels high by 475 pixels wide, banners didn't take up much space on the screen and they could be made to load fast. They were almost exclusively totally cross-platform GIF files. They were just like another picture on the page and they could be links. Publishers were quick to adopt another standard beside size on the page. You can create a graphic so rich in that little space it would take minutes to download so publishers added a file size limitation too. The actual size requirement depends on the individual site/publisher but it's never more than 10K.
Just a putting static image on the page, e.g. a picture of a product or its symbol, can increase brand awareness but it can't add to the interactive web experience or deliver much of a message. If the banner were a link the viewer had an interactive choice; they could click on it and learn more about the product or the company.
A link made advertising even more attractive because companies were not just creating brand awareness they were engaging their potential customers. Customers were coming to their sites to find out more. If there was nothing at the site, if it had no content or fun, people quickly went back where they came from. If the viewer perceived no reason to stay at the site, not to buy or learn, they were gone. It was not perfect but it worked.
Publishers had finally found a way to make money from their web sites. Advertisers had found a way to tap into another medium and if not a new group of potential customers, at least a way to reach people again. To determine prices for the space the publishers again went to a traditional advertising model. They would charge according to how many people saw the ad. It's called CPM, Cost Per M (thousand) of impressions or page views and the price advertisers would pay was determined by how many people saw the page their ad appeared on. The advertisers had an equally specific way to measure the success of their banners, click-through. A click through occurs when a viewer sees a banner ad and clicks on it to go to the advertiser's site. The number of click throughs that occur, divided by the number of times the ad appeared (page impressions) would determine the success of the banner. As the Internet/WWW grew so did the number of page impressions (Yahoo! Has 40 million users worldwide and can measure page views in the billions per month) and the number of click throughs but CPM for banner ads has not risen with the numbers. In fact, it's the other way around; CPM in the Internet world is eroding.
WHY BANNER ADS GET MORE FOR LESS
For the first time in advertising the numbers of impressions can't be padded with the people who left for the bathroom during a TV commercial or those that left the television on when they went to the store. For magazine's it's the same story. The CPM rate for a magazine is based on the circulation regardless of whether the entire potential audience sees that ad or not. This means that with the same stated audience sizes TV and print businesses have more ad impressions for sale than most Web sites could ever hope for.
If a magazine has a circulation of 100,000 and 100 advertisers the magazine bases its ad prices on distribution to all 100,000 of its readers. If you multiply that by the number of advertisers it would have 10 million impressions for sale each month.
A Web site doesn't base its success on the number of advertisers it has but on the number of people who actually visit the site. Also unlike the magazine it can't simply increase the number of pages it has. The Web site has only the pages the viewers generate themselves with browser requests. To figure out the number of ad impressions a Web site has for sale it needs to know the number of visitors it has and also how many pages each unique visitor looks at each month. The online research firm Relevant-Knowledge analyzed its 50 most-trafficked sites and found the average user visited 17.2 pages per site per month. These numbers were skewed somewhat by sites like eBay.com (172 average page views per unique visitor), Sportszone (101) and Hotmail (87). Without these 3 sites the remaining 47 averaged 14.7 pages per visitor per month.
If a site has 100,000 unique visitors per month and each visitor averages 15 pages per month it has 1.5 million ad impressions to sell each month. Even if the publisher stacked 3 ads per page it would only have 4.5 million impressions, less than half the number of impressions the magazine has with the same size audience. To earn identical revenue the site would have to charge twice as much. This makes it very hard for the small niche publisher to make money. The big sites, the portals and search engines, are doing fine. Ted Risdall of Risdall, Linnehan Interactive speaks for anyone placing ads on the big sites, "Yahoo! Won't budge on their Rate Card." Yet Yahoo! Is one of the few sites in the top 50 who are reporting a small profit from advertising revenues?
Bryan Wiener, Director of business Development at Standard & Poor's Personal Wealth site (www.personalwealth.com) says. "I'm not sure where the cutoff point is, but I think you'd need 10, 20, 25 million page views a month if advertising is your only revenue source, and then you're still probably not going to be profitable." NetRatings, a traffic reporting company, says there are fewer than 100 sites that have more than 10 million page views per month. The picture for advertising isn't as lucrative as it appears from the number of banners you see.
There are too many banners for sale, there are too few sites where advertisers can reach a large audience and people just aren't clicking through any more. Viewer response to standard GIF banner ads has fallen steadily.
Here are some actual numbers from NetRatings for an average user during the week of December 6-12, 1998:
If this is the biggest Holiday season ever on the Web, the click through rate for ads on these pages .062% doesn't reflect it. Less than 1% of the banners seen were clicked on. It make CPM at any price seem high. Click through is more important for some companies than others. If you run a mail order catalogue business and you need people to come to your site to buy, click through is very important. If you are a soft drink company and you just want people to see a banner saying, "Drink More Burpo!" click through isn't. But the numbers show that people who click on a banner have better brand retention than those who don't. If they clicked on your banner they remember you at the supermarket more often.
WHY DON'T BANNERS WORK?
But advertisers want you to buy their products and they have ways to attract your attention. The Internet is nothing if not constantly evolving. Remember when a part of its promise was to deliver a multimedia experience to your desktop. Nothing would please advertisers more. They know there are better ways of getting your attention than a simple GIF animation loop. They know that the Internet audience is a willingly interactive group of people. They know that the more appealing they make their product look the more likely you'll be to remember them come purchase time. They know those things and there are a whole bunch of developers and programmers out there with products and plans that can give it to them, and us. It's called Rich Media (see sidebar) and it's on its way.
WHY RICH MEDIA?
The first evolution of rich media (if you exclude the GIF animation) was probably the HTML banner. HTML banners were a significant step above static GIFs. They are a seperate HTML page within an HTML page, sort of like a frame. In fact they can be just like a frame in that the page you're viewing can be on one server and what you're seeing in the banner can reside on another. HTML banners can include sound, animation and interactivity within the banner. This makes it ideal for contests and games, like kicking a football through the goal post with a mouse point and click. If it goes through the goal post, the banner moves you to the sponsor's site and you register for a promotional drawing. You've had some fun and the sponsor has generated a solid lead. Good for the viewer, good for the advertiser.
But there are trade-offs, the primary one being slower loading times. While most GIF banners are under 10K, the average HTML banner is about 35K. The majority of HTML banners are based on Java program files. Sites that accept interactive HTML banner ads will do so only for a premium. They charge advertisers more because they use more bandwidth and because they need time to test them before placing them on their pages.
Advertisers love them, and pretty much every other kind of rich media ad, because they are more effective, plain and simple. One of the first studies on rich media proved it. In 1997 i-Frontier designed a rich media campaign for Atlas Editions' Star Trek Universe trading cards. Click through was 4 to 8 times higher for their HTML banners than their static GIF banners. The actual rates were 2%-4% versus .5%-. 8% for different versions of the same banners.
HTML is the bottom end of the rich media heap. Even though the majority of them use Java or Javascript applications. Above them in the pile are other forms of interactive banners based on Java applications. While most of these are loved by advertisers some of them are literally hated by both the site owners and the people who encounter them. The most denigrated of these banners is the pop-up. Pop-ups appear in a new window, new screen or are otherwise separate from the media in which they are contained. Simply put, you click on a link, go to a new page and suddenly another browser window appears with the ad on it. Usually these new windows are smaller than the page you actually requested but they always appear on top of the page. The most commonly seen is the window that appears every time you click a link to a page residing on a free homepage GeoCities, Angelfire or Tripod site. These sites offer free home page space to millions of users and they require the HTML code for the pop-up appear in the code of every home page. Pop-ups can be programmed to close on their own or the viewer can choose to close them at any time. The interstitial pop-up is the most commonly used pop-up. It is used to advertise while users wait for downloads. More about interstitials later.
Another class of HTML ads are called rollover driven ads. They occur when the cursor moves over a particular part of the screen. Everyone who has used "Balloon help" has used rollover programming.
Because these HTML ads are most often powered by either Java or JavaScript programming they can contain any multi-media and/or interactive elements the creators want to employ. Even without these elements, HTML ads are more effective than static GIF banners and produce higher click through rates.
An interesting hybrid approach has been developed by a company called 9th Square. Their solution is called an E*Banner. 9th Square says it's a way of taking the advertiser's site to the consumer instead of trying to get consumers to come to their site. Here's how the E*Banner works.
You call up a page. Above the content, as usual, is a banner like any other. You click the banner and it expands instantly to cover about a third of the page. The banner contains a bit more information about the advertiser and, say, a button to click for more information. When you click it the ad changes to a form asking for your zip code. Click it again and the page delivers a list of the five closest dealers. You write down an address or phone number, collapse the banner and you're right there on the page you left. As far as viewers are concerned they've never left the site they wanted in the first place. No confusion about where they are, no series of back clicks to return to where they wanted to be when they began, yet they've also done everything they wanted to do at the advertiser's site too.
The E*Banner s a good solution that addresses the interruption factor. There is another group of ads that has taken another approach Flash or Shockwave ads are more aggressive in their approach. Both Flash and Shockwave are products of Macromedia, Inc. Macromedia has been a leader in the creation of PC/Mac based interactive, animation and audio software products since the PC was invented. Their Director program and Lingo authoring language are industry standards in the Interactive CD-ROM industry. Shockwave is probably the most popular program on the web for delivering interactive games over the 'Net. Flash is capable of delivering high quality vector based graphics, illustrations and animated web pages. Exciting and fun they both suffer from the same Internet curse, large file size.
But until recently they both were under a much darker cloud, viewers had to have a player/plug-in on their machine to make them work. If you didn't have the plug-in a click on a Flash banner would bring a dialogue box asking you whether you wanted to get the plug-in you needed. A Yes response meant a trip to another site; a download, install and a computer re-boot to make the plug-in active. Not many people were willing to do that just to see Mickey dance. But there's no question once you had it, it was worth it. The sophistication both programs can bring to the web is stunningly attractive and engaging. Interest in the sponsor's product is definitely more positive when they supply an exciting or entertaining experience; Flash and Shockwave can do both. The player/plug-in problem has been eliminated too for many users. They are both installed as a part of Microsoft's new Windows98 operating system. For users coming to the web on new computers, they will never know they weren't always a part of Explorer.
Another technology that as been slow to be accepted is streaming audio and video. While the industry leader RealAudio has worked hard to apply their expertise to content sites like NetRadio and CNN, their use in advertising has been minimal. Again it is the need for a player on the user end that has stopped advertisers from employing them in their banners. Why use a medium a vast majority of the audience won't have to deliver your message. At the same time, they are hungrily awaiting it's development. Moving visual images are much more effective at grabbing and holding a viewer's attention. They are also much more effective in getting a viewer to interact with a banner. When the Sci-Fi channel used a series of "live" video banners their click-through rate increased tenfold. They also found that visitors who did visit the video streaming sites stayed an average of four times longer than at sites that did not use video. The same has held true for other users of video in their Web advertising.
The use of video is attractive to advertisers because it directly impacts one of the major complaints advertisers have with banners, the limited message they deliver. In the same sized space, streaming video with or without accompanying audio can deliver a much more detailed or entertaining experience for the viewer. If that is true then click through may become unnecessary, thus making the ads more attractive to viewers. In the end this could please publishers as well, if VBanner ads are more effective they should be able to charge a greater CPM for VBanners.
INTERVU, is the industry leader in creating banners using looped video. In fact they hold two patents for the delivery of video and audio over the Internet. The other edge INTERVU has over encroachment of its VBanner dominance is their network. INTERVU's proprietary network infrastructure utilizes a configuration of delivery centers intelligently and strategically distributed across ten Internet backbones. This distributed network is effectively INTERVU's operating system. It load balances to optimize audio and video file availability for Web sites and users. The INTERVU Network can determine which delivery center is electronically closest to the end-user in order to quickly and efficiently deliver the multimedia from that location, thus avoiding data bottlenecks and backbone failures and ensuring high quality delivery. The INTERVU combination of production expertise and their efficient delivery system has created VBanners with enviable click through rates for many major corporations.
The click through on VBanners is incredible when viewed against static banners. A golf supply company, Goldwin Golf, ran static banners on GOLFonline. They're click through rate was 1.8%. They went to INTERVU and produced a banner with a 3 second video loop of Nick Price swinging a Goldwin driver. Then they ran the V-Banner on the same site, GOLFonline, with the same frequency as their old banner for six weeks. Their click-through rate went up to 5.6%-5.9%, an increase of 250%. A second VBanner running in the same spot had a similar response and averaged 5.8% click through. Overall the VBanner generated a 10 times increase in the request for Goldwin catalogs.
Budweiser also went to INTERVU for a VBanner. The Bud VBanner ran on the Lycos search engine for about four weeks coupled with the keywords "bud", "budweiser" and "Anheuser-Busch". The average click through rate was 42%. Now the placement and keyword purchase certainly had something to do with thee rate. People requesting a search for those key words were more than probably looking for information contained within the banner.
Fusion Films ran a VBanner with a teaser clip from their "Nobody's Heroes" snowboarding videotape. The click through went to an eCommerce page where users can purchase copies of the video in the VBanner and others. The click though rate for the VBanner was almost 40%. The attraction of VBanners is overwhelming for advertisers because it's like TV. It gets peoples attention and it gets them to act. The problems again are the same, file size and viewers/players/plug-ins. But the evidence of their effectiveness is overwhelmingly in their favor. It is only a matter of time before we see them on more and more sites.
The final category of rich media banners aren't always banners. In fact, they can be entire full screen web pages containing any or all of the production technologies of other banners. What make these ads different is not so much the technology used to produce and view them as it is the way in which they are delivered. As a group they are called interstitials. Any ad in fact becomes an interstitial when it runs while the user is downloading another page. In that way the GeoCities banner that pops up when you request a GeoCities home page is an interstitial.
An interstitial ad is far more intrusive than other forms of online advertising. When you click on a link at a site that is using interstitial ads, you do not jump right to your destination. Instead, first a full page advertisement appears and several seconds after that page finishes loading your browser will display the page you wanted.
Advertisers love interstitials because they so closely emulate TV and print ads. If the user wants to see the Web page they requested, they have to watch the ad. Just like if you want to watch the Movie of the Week you've got to watch the commercials. Other benefits of interstitials include the level of information they deliver and they have proven to be one of the most successful of all ways to advertise on the Web.
Interstitials although they are indeed delivered from server to client like Web pages, are not delivered from client to user like most Web pages. Interstitials are loaded in the background, the content of the ad resides on the users hard drive until a web page calls for the ad and then it plays. While the concept is very promising the execution is even more difficult than with other forms of online advertising because the user must have "client" like software installed to trigger the ad to play. Getting users to download and install the software presents the same problems as plug-ins. User motivation to download and install software intended solely for playing ads is negligible.
While there are more and more interstitial ads running on the net (General Motors Corp ran an interstitial on Real-Networks Daily Breifing news site that garnered a 6.3% click through rate) the most successful interstitial implementation today has been done by Berkeley Systems. Berkeley Systems proprietary Interstitial system was designed expressly for the company's net shows, and there are no plans for licensing it at this time. This high-performance sound and animation display system was derived from technology developed and modified from years of writing screen savers (boy, have those toasters grown up!).
The interstitial ads that appear during "You Don't Know Jack, the netshow" and its companions are authored in Macromedia Director (as is Jack). Then, according to Jack Eastman, Chief Technical Officer at Berkeley, "A proprietary tool extracts the art and screen choreography data from Director and emits it in our own highly compressed graphics format." Music is compressed using Xing's MPEG audio compression (speech is compressed at a much higher ratio using the DSP Group's TrueSpeech 6.3 kbit algorithm).
All of the media elements are fully downloaded before they begin playing, being scheduled to arrive just prior to their use. This delivery schedule is carefully tuned to minimize the initial download while making sure that few users see any interruption of their gaming experience while online. The entire game engine is delivered in the initial download, and there are no extra components such as Shockwave or DirectX that the user must already have. This makes it both immediately accessible to the user, and able to run on a plain-vanilla 486/66 PC. The game is launched by clicking a button on the web page: the installation software having registered the Netshow as a "helper" application in the browser and running then in a similar way as if, say, RealAudio were called into play, being outside of the browser itself.
In 1997 Berkeley Systems commissioned a study that found interstitial ads that appeared during breaks in their online games - generated higher brand recall than ubiquitous online banner advertising and were even more effective than traditional ads on TV and in print media.
Berkeley is running interstitial advertising on its beZerk online gaming site, which hosts the online version of Berkeley's popular "You Don't Know Jack" trivia game. MBInteractive, which conducted the study, found that consumers exposed to a single interstitial advertisement were 64 percent more likely to recall seeing advertising for a specific brand; on average, consumers exposed to traditional banner advertising only experience a 30 percent increase. With interstitial ads, awareness of tested brands increased an average of 7 percent, with newer brands seeing greater lift than well-known brands, and, on average, consumer purchase interest in the advertised brands increased 14 percent.
The study also concluded that interstitial advertisements are twice as effective as traditional online banners and three times as effective as television advertisements. Because the five to ten second interstitial ads are analogous to TV commercial breaks, and because "You Don't Know Jack" appeals to a more mainstream crowd than most online games do. Berkeley has been able to attract big consumer goods advertisers like Jack Daniels, Nabisco, and Hugo Boss; most sites make do with ads targeted solely at gamers.
The study may well spark an interest in online advertising by mainstream corporations. The cost of developing an electronic ad and the lack of a standard ad format have kept many advertisers out of the market. Game developers, for their part, have been trying to come up with creative ways to integrate advertising into the games without disrupting play. Interstitial ads require that the game be designed with periodic breaks in the action.
SO WHERE IS IT?
The Wall Street Journal Interactive Edition and CNN.com will accept HTML rich media banners but are reluctant to run other formats such as video citing slow download times and integration with third party servers as its reasons. (Often streaming video or audio ads reside on a third party server at an ad serving or streaming media company.) Turner Interactive Sales will not accept certain types of rich media ads such as interstitials and pop-ups. But these attitudes are changing.
Ad networks that welcome rich media are growing every day. San Francisco based Flycast Communications Corp. has over 500 sites in its network and approximately 200 of them are running rich media ads. The DoubleClick network has 170 sites that consider each rich media request on a case by case basis. Jamie Byrne, head of online marketing and sponsorship development has said, "We need to be able to accept any and all types of enhanced media." Ad network 24/7 Media has over 100 sites in their network and also reviews rich media requests on a case by case basis. LinkExchange, with over 250,000 sites in their network, is accepting enhanced media banners once they have been tested.
The biggest sites, like the search engines and portals are still lagging behind. Once, and a very short time ago, they were all aggressive innovators. It's what took them to the top. Today they have taken on the conservative "wait and see" attitude of traditional big business. Eventually, as rich media technologies standardize and bandwidth increases, they'll be the ones touting their rich media capabilities the loudest. Until then it's up to the advertisers, the agencies and designers who create banners, and us, the everyday Internet users, to keep the pressure on the publishers and site owners to give us what the Web promised; an exciting, stimulating, interactive, multimedia experience. Even if it does have commercials.
iBiz |
Whether we like it or not the Internet is now about money.