In 2011, I gave a talk to a group of online ad-sales people who were so full of the multitude of different ways that they could target and quantify their product, they literally no longer understood what brand advertising is, or why it exists, or why brands would be so foolish as to spend so much money on it. They’re quants, living in a world where something only has value insofar as it can be quantified, and where the unquantifiable therefore is perceived to have no value at all. In other words, they’re basically in the direct-marketing business: they’re the digital version of junk mail. As a result, just about every website in the world is in the business of delivering that digital junk mail to our computers and iPhones and iPads.

This, then, is the biggest reason why TV ad dollars are not going to become online ad dollars: online ads simply don’t do what TV ads do. TV ads are large and beautifully produced and expensive, and they’re presented on a beautiful screen without distractions: they fill up the screen, and 30 seconds of time, and they appear often enough that they become part of the world of the people watching 145 hours of TV every month. Online ads don’t behave like that at all: they’re easy to ignore, there’s nothing inherently interesting about them, and insofar as they grab your attention, they tend to do so in a very annoying way, by preventing you from reading or watching the thing you were looking for.

Hence the rise of so-called native ads: things you want to read and look at and click on. There’s a certain amount of promise there, and the native-ad industry is certainly going to grow from its present size. But it’s tough: building these things is a huge amount of work for the advertiser, with no guaranteed payoff. And selling them is even more work for any publisher.

And here’s the next big problem with selling online advertising, especially native advertising: it’s really expensive to do so. While online journalism is still cheap, online ad-sales staffers tend to cost a fortune, especially if they have a clue what they’re doing. This is something the Meekers of the world would do well to remember: the ad dollars spent online are spread across so many sites that a massive proportion of them end up just going straight into the pockets of the people selling those units, or else to the various ad networks and other intermediaries which have popped up in a very busy and messy space.

This kind of thing just doesn’t exist in TV or even in brand advertising more generally — areas which are much simpler, much easier to navigate, and which sit much more comfortably within consumers’ comfort zone. And it’s not going away. I was told this evening that Buzzfeed alone has no fewer than sixty ad-sales people, all of whom are out there, knocking on doors, taking potential clients out to lunch, and generating income one hard-won deal at a time. That doesn’t scale.

Indeed, if you want to get your brand out there on the internet, you can try buying ads on websites, or you can try going native on a site like Buzzfeed, but the fact is that the whole point of the internet is that it disintermediates: it’s great at drawing direct connections. Hence the rise of what’s known as “content marketing”: why buy ad space from a publisher, when you can be the publisher instead? We’re still in the early days of this, but already musicians are discovering that brands are much friendlier — and pay much higher rates — than record labels, while American Express has been employing extremely good journalists for years.

On top of that, as Liz Gannes and Noah Brier note, nobody “goes online” any more: the internet is becoming an ambient background thing-that’s-always-there, rather than a mass communications medium that people consciously think of themselves as paying attention to. When you pick up a magazine, you do so because you want to read it; similarly, when you turn on the TV. But the internet is different: your phone is always just sitting there, and sometimes it beeps at you; your computer is always on your desk at work, and it’s never not online. In a mobile world, the distinction between being online and not being online is an increasingly silly one to draw. And as a result, the idea of using “time spent online” as a useful metric of anything, really, is equally silly.

Felix Salmon is an Audit contributor. He's also the finance blogger for Reuters; this post can also be found at