If there is one recurring theme as it relates to media, technology and advertising it is that, by the time an advertising model is well defined and universally adopted, a new technological advancement comes along and ruins it. Radio was the beginning of the end for print. Television was the beginning of the end for radio and the Internet is here to kill them all.

In the early days, advertising was really incredibly simple:
“This product does X. Here it is. If you want it, it costs Y.” The most important thing was to let people know the product existed and add the most basic information about how and where they could get it. A sign outside your shop on the street did the job.
Then as competition started to grow, vendors had to not only inform their potential customers about a specific product but also let them know in what way this product was different or better than all the others that were out there. They also had to reach as many people as possible. From signs in front of their shops, vendors moved to signs that were placed higher up, so that more people could see them and then to newspapers where even more people could see them and so on. You get the idea: “follow the audience”. And the audience usually follows the newest shiny object.
From print to radio then on to television, at each stage advertising evolved with the technology but it wasn’t until the radio era came along that the business model changed from “I will make some extra money by selling some ads” to “my business is entirely relying on ads to survive”.

Since radio was free for everyone, there was no fee charged for listening to it but there were plenty of expenses that came along with it and radio broadcasters took advantage of the fact that they could reach a lot of people to make money. In more than one way, as the Payola cases describe. http://en.wikipedia.org/wiki/Payola
The advertising-based model then was adopted by broadcast television as well: since there was no easy way to keep people from accessing the airwaves or charge them for it, they would have to watch and listen to ads in order to view the content. This model became so prevalent that most content providers and advertisers started taking it for granted; claiming that consuming the content without watching the commercials is theft.

One of the first “no ad skipping” cases in the long line of fighting technology advancements in the courts was the ReplayTV case in which the then-CEO of Turner famously said that “Any time you skip a commercial . . . you’re actually stealing” but that “there’s a certain amount of tolerance for bathroom breaks.”
In other words, you’re allowed to go to the bathroom during the commercial break but that’s it.

Oh, and they will also pump up the volume to deafening levels so that you will, for sure, still hear the commercials even if you can’t see them.  It’s the law.

More on that case here:
http://writ.news.findlaw.com/commentary/20020509_sprigman.html

This was all before the Internet came along and changed the industry yet again.
What the Internet allowed people to do was switch the advertising process from being a monologue and one-way communication to a (sort of) dialog. Digital publishers knew, or claimed to know, whenever someone actually looked at an ad and so a new business model emerged: you only pay for what you buy.

“You only pay for the ads that someone has seen” was the famous tag line everyone was using at the time. You don’t pay for any impressions that were not displayed like you would in print, radio or television. This was huge! No more wasted money. The other major advantage was that you could also get an instant response from people in the form of clicks.
This opened up a whole new world of possibilities in terms of both crafting the message and getting it to the right audience, at the right time, in the right context.

On the Internet, they say, nobody knows that you’re a dog. But that may be the only thing “they” don’t know about you. Unlike in print, radio or television, on the Internet we all leave a trail of bits that can be used to learn more about us, what we do and, ultimately, what we want.
So with the advent of this new medium, the issue became not only getting the message to as many people as possible but also, so as to not waste money on people who couldn’t care less about your mortgage refinancing or belly fat product, you could now show your ad only to people who are likely to buy your product.

Does this sound scary? That’s because it is. However, remember what I said earlier that nobody knows you’re a dog? All this information about you is inferred from everything else that you’ve done online. So even if Visa or Pepsi or Amazon know exactly who you are, where you live and how much money you make, they have to pretend that they don’t. For privacy reasons.
Privacy is one of the biggest issues that interactive advertising is grappling with. Publishers and advertisers constantly have to walk the fine line between not wasting any money on a dog and admitting that they know you’re a dog. You have the right to be a dog on the Internet if you so choose but unless you give Amazon or Google your name and address (first-party data is what they would call you) AND explicitly authorize them to use your name and address for marketing purposes (opt-in is the process by which you would allow this), they are not allowed to tell anyone that you “First Name, Last Name” are a dog. So what they would typically do is, they would create a category called “Dog” and put you, as an anonymous random number, in that category and tell their advertisers that they have 250,000 dogs that they can reach on their sites.

So the new digital medium created a whole new world of possibilities in terms of targeting: geo-targeting (where you are), demographic targeting (who you are) and behavioral targeting (what you do) are just some of the new methods of getting the message to only those who are likely to be interested in it.

On the other end of this pipe we call “Internet” is measurement.
In print, radio or television there are ways to estimate the effectiveness of an ad but they will always only be estimates. The internet, because it is a two-way communications medium, allows for instant response from people which lead to a whole range of new pricing methods that evolved from the traditional cost per thousand (you pay for every thousand ads displayed) to the more sophisticated cost per view (you only pay if someone watched the entire ad), cost per click (you only pay if someone clicks on the ad) , cost per transaction (you only pay if someone signed up on your website) or cost per acquisition (you only pay if someone actually bought your product). These are performance-based buys and are attractive to advertisers because it is tempting to only pay for what works.

That is not to say that the older models didn’t have advantages that got lost in the shuffle of the new technologies.

With a print ad, you may pay for 300,000 copies of your ad that may or may not be seen by someone but, at the same time, your ad will sit there, in a doctor’s office or on someone’s coffee table for months or even years and there may be a new buyer who happens to pick it up at some point when you don’t even expect it.

With TV, there’s no question that the impact is still huge and people will remember that ad that they saw during the Super Bowl maybe even for years to come.
Perhaps the most famous example is the ad for the Apple Macintosh commercial from 1984.
https://www.youtube.com/watch?v=VtvjbmoDx-I

It only aired once on national television and 30 years later people are still talking about it.
That sort of impact tells us two things:

Digital advertising is up against some pretty steep competition in terms of lasting impact, brand recall and instant reach
Sometimes the message is the message.