With TV advertisers facing challenging economic times, Underscore Marketing's president offers an online alternative.

When Bob Garfield explored the notion of what might happen if the broadcast model collapsed -- even partially, before the digital marketplace was able to meet the subsequent demand for inventory -- his prediction was dead on. If the television business collapsed tomorrow, there's no way digital could absorb more than $40 billion in spending.

Looking solely at online video as a potential replacement would today be silly. It simply can't reach the number of households television can. The top video site on the web doesn't even have a pre-roll ad model, so even if advertisers wanted to use it to advertise, they would have to use something other than a 15-second or a 30-second spot.

Still, if certain advertisers wanted to replace their television schedules with online video, I have no doubt it could be done, and it might end up being more efficient for them.

It all depends on the incidence of use for a product or service. A beer brand and a prescription drug brand might both advertise on television, but it's highly unlikely that more people have the condition addressed by the prescription drug than drink beer. A higher percentage of the prescription drug ads thus go to waste.

It's these kinds of lower-incidence advertisers that could take advantage of online video in a big way. As I've discussed in this column before, demographics are a really crummy way of predicting purchase behavior. Behavioral and contextual targeting is the way to go for these advertisers, as it can cut down on waste and make one ad impression do the work of 10 or 100.

Broadband Enterprises, a leading video network, uses a contextual engine to deliver relevant video advertising. It can reach over a third of the total online video audience with tens of millions of monthly uniques. At that kind of scale, a contextually filtered buy could reach a huge number of potential users of a lower-incidence product in a very efficient way.

That's just one network. Think about what happens as other sites and networks with targeting tools apply them to video inventory. AOL is a top 10 aggregator by itself. How long before the Tacoda technology AOL acquired can deliver behaviorally targeted video advertising at scale? What about when Yahoo, another top 10 video aggregator, applies complex profile information and/or purchase behavior to video ad targeting?

The point here is that when online video takes a bite out of television it will do so because video advertisers see value in targeted advertising, not because online video viewership begins to approach that of television.

The technology is certainly here today. If we have any issues, they're tied to integration of these targeting technologies, which up until now have been used to deliver targeted banners, so that they can deliver video inventory. That's not a huge challenge, considering what is at stake here.

For lower incidence products, we might soon see sizeable online video schedules -- targeted contextually and/or behaviorally and frequency-capped -- taking serious money away from television.

The two key variables here are scale and targeting criteria. The scale will be here soon, considering online video grows by leaps and bounds while TV viewership declines a couple percentage points per year.

So, it's time for forward-thinking advertisers to begin testing targeting criteria. Different video sites and networks will have different targeting capabilities, so testing is important if a brand is to take advantage of the medium in the near term.

Get started now.

Tom Hespos is the president of Underscore Marketing and blogs at Hespos.com. Read full bio.