First there were hyperlinks. Next there were buttons. Then we had tiles. Following that there were banners. Then pay-per-click search.

All of these have served, for a long time, as the meat and potatoes of online advertising units. At first, costs for this advertising were flat rates for placement. Kind of like slotting fees for shelf space, advertisers paid for real estate and kept it for the term of the contract worked out between marketer and publisher.

Not long after, the impression became the coin of the realm for media, the currency of exchange for advertising inventory.

But over the years, a lot of other ad formats have come online. We�Š�e got large display ads that take prominent positions on a web page or are served in the transition of clicking from one page to another. There is streaming audio that is served either as part of a banner; or it is served either before or after an audio contents stream. And there is video.
�
What all of these formats, more in demand and in wider use than ever before, have in common is that they are all formats that have originated offline, over traditional media channels.

Online now offers advertisers the opportunity to use -- in one medium -- many of the same tools and strategies developed for various other media. The web has more or less become an environment of convergence.

Broadband has brought to online the ability to generate and serve an increasingly complex array of advertising formats. And broadband�Š� growing ubiquity has given advertisers access to larger audiences, making the development and deployment of these more robust ad formats an efficient and effective proposition.

But how do we treat these formats? How do we plan and buy them? Do the old formats conform easily to online ad models, or do they bring with them the methods for planning and buying that they inherited from their forebears?

Like Print

There�Š� nothing all that new about large-sized display ads online. They�Š�e been around in various forms for years. We�Š�e had interstitials and superstitials; we�Š�e got quarter-page and half-page units.

With a larger data pipe now servicing just over half of the online population, more can be done with these units than has heretofore been possible.

But should we simply treat these as if they were just like bigger banners, selling them at a slightly higher premium, and leave it at that? How do I, as a buyer, determine their actual value versus their value relative to the majority of banner-based display advertising that is found on the web?

I�ŠŽ like to propose that these kinds of ad units be treated differently. I suggest that ad units such as these and the media they run with be dealt with much in the same way as we deal with print.

Not all online ad vehicles can be treated like print, and for the moment I'm going to set those that can�Š� aside for a different kind of evaluation. (I'll get back to them shortly.)

For a property that behaves a lot like print (e.g., Salon.com, special-interest content sites, financial news and newspaper sites, et cetera), online inventory should be sold much like print inventory: a fixed position that exists on a page for the life of the issue.

When I buy print, I pay a CPM derived from a guaranteed rate base. A rate base is a minimum circulation for a magazine guaranteed by the publisher. It is a way to guarantee advertisers that at least some fixed number of people will have a chance to see their advertisements. It serves to fix an "opportunity to see" (OTS) and set a floor for communication delivery.

Some �…�?number of unique individuals will have an opportunity to see my ad. If I buy placement on The Onion in the July 13 issue, the ad stays up for the week. The ad unit needs to be permanent. When I "turn the page" on an article, maybe there's another advertiser there who has a fixed placement. CPMs for circulation can go up some, but the numbers of impressions become less of a focus, and the problems of remnant inventory, for those sites that have this problem, can go away.

Yes, impression CPMs will effectively go down, but that's okay. If, as an advertiser, I buy space in an online publication and get more impressions for my dollar, the efficiency of that effective CPM is increased, and I will be more likely to buy in that publication again as a result.

What happens then is that out of pocket for an ad placement goes up, the CPM goes down, and inventory management becomes a simple matter of availability of ad space as it relates to actual audience rather than a monthly frenzy of booking millions of impressions.

Also, clients can more easily satisfy their need to see their ad where the agency says the ad is running. The client�Š� ad, instead of being in a dynamic rotation with dozens of other advertisers, will be on the page that has been negotiated for the time it takes to deliver on the guaranteed audience.

Like Radio

Streaming audio content is now a given online. Though there are still the occasional 揵uffering?issues, most online audio content now serves as smooth as an early Spring run-off.

There are a lot of people who now stream audio content on a regular basis, be it online radio or the more talked about podcast.

According to @plan�Š� Summer 2005 release, there are over 34 million people who have listened to the radio online or have streamed audio content in the course of a month; nearly 12 million of them do so every day.

Arbitron sets the figure higher, reporting monthly internet radio audiences at an estimated 37 million Americans. According to data found in the Arbitron/Edison Media Research �Ÿhe Internet and Multimedia 2005: The On-Demand Media consumer, as of January 2005, 15 percent of Americans say they have listened to internet radio in the last month, and eight percent have listened to internet radio in the past week.

Whichever set of data you accept, in either case the numbers are big enough that they should be attracting advertisers?attention.

So why aren't more advertisers getting onboard?

One of the reasons is that those selling streaming audio are having a hard time finding the right person at an agency to approach. Secondly, it isn�Š� uncommon that no one at the agency is sure who the right person is, either.

And finally, no one is sure of the best way to position, buy and sell streaming audio. Impressions? Spots? Gross rating points (GRPs)?
 
Streaming audio as an advertising vehicle needs to be made comprehensible to media people and brand managers who already understand radio.

One solution might be to treat streaming audio like it is radio -- a midday radio daypart, to be exact. When buying radio schedules, buyers have four main dayparts they can buy: a.m. drive, midday, p.m. drive, and evening/weekends.

The midday daypart used to be a way to supplement your radio schedule and get messages into the ears of working people listening to the radio in the office. But with streaming audio, advertisers trying to reach any demographic or psychographic can get their messages out to their audiences not only locally -- as most radio is bought -- but nationally, because they can buy the specific genre or format only and not a whole radio schedule. And again, they can speak to their audiences at a time when no one else is (except maybe their bosses).

Next thing is to determine inventory units. A per-spot or GRP could serve as the currency. For formats that have a great deal of listeners, it is easy to determine what the GRPs would be and how an advertiser's online radio schedule fits the overall communication impact of an offline radio schedule. Advertisers could use online audio-streaming buys as a "midday daypart" to supplement their offline radio buys.

Doing this accurately and effectively will require greater research on the various online audio outlets, but this is being worked out. Arbitron, in conjunction with comScore and Scarborough USA, have a great deal of data about online audio/radio audiences and are continually working to get more precise, granular data.

Like TV

So much as been said about video online in the last few months that sometimes I think I�ŠŽ rather chew aluminum foil than read more about it. I even wrote a column dedicated entirely to video just a few weeks ago. But let me just say a few words about video online in this context.

Video is going to become the preeminent online ad format in the months and years to come. But questions about how it is to be evaluated alongside the rest of the online media communications package remains to be determined definitively.

I�ŠŽ like to suggest that, like television, online video�Š� communication delivery be evaluated against audience delivery rather than impression delivery. Before the eye rolling begins and protestations abound about using old media metrics to assess a new media execution, think about this.

As I am wont to say, people make purchase decisions, not impressions. If I�Š� a client with a product to sell, being able to evaluate my media in terms of the potential number of people I am going to reach with that media is a lot more important than the number of impressions I am going to deliver against that media. Some number (X) of people is going to be exposed to my message. Some number (Y) of people within that X is going to be persuaded by that message. And some (Z) number of people within that Y is going to act on that persuasion. From this data, I can determine the success or failure of an advertising campaign and calculate projected number of units to be sold and how to distribute them.

One of the unique things about the internet is that it does, after all, act as a convergence medium. Some online media vehicles act like print, some like broadcast. To think of online monolithically in terms of how it delivers media, what media currency is used, and how that currency is evaluated is to ignore the variety of online�Š� content delivery possibilities.