Mention Pandora to a traditional over-the-air radio person and you’re likely to get an earful.

They may or may not listen to the online music service, but chances are they’re intimately familiar with its recent initial public offering and even more familiar with all the arguments as to why Pandora just can’t succeed as a business.

For sure, there’s lot of debate over Pandora’s business model, and Pandora period. (“Pandora IPO at $16 a share - Biggest wonder or tech blunder?” read a headline after its IPO two weeks ago.)

But among mainstream radio people, the doubts over Pandora are tinged with defensiveness and real hostility among many.

It’s not hard to understand why.

Certainly part of it is the service’s huge and rapid growth; it now claims some 94 million subscribers, and new listeners sign up daily. And of course Pandora is taking some advertising from terrestrial radio.

But the bigger issue is the threat Pandora represents to traditional radio’s own business model.

That model is based on the longstanding agreement that exists between broadcasters and their audiences, the quid pro quo. The broadcasters provide hours of entertainment and information in exchange for listeners putting up with a slew of commercials, anywhere between 12 and 18 minutes an hour.

Pandora offers listeners a radically different contract: far fewer commercials per hour, a couple of minutes at most, but the same or greater amount of entertainment value, since listeners can program their own music choices.

That very different agreement explains why Pandora has seen such growth, and it represents a real fix for traditional radio. To compete head to head, it would have to cut its commercial load substantially, but that would be financially impossible.

Its only recourse is to talk up all the benefits terrestrial radio provides that Pandora cannot, such as its close and longstanding ties to local communities, both listeners and advertisers. But while those ties are quite genuine, it’s highly debatable as to whether they’re strong enough to halt or even slow the flow of listeners to Pandora.

Another option for terrestrial radio would be to steal Pandora’s thunder by offering similar services, and indeed Clear Channel is beefing up its iHeartradio to directly challenge Pandora with Pandora-like features. A beta version is set to roll out this fall.

But broadcasters have stopped short of an all-out attempt to invade Pandora’s space for a very simple reason; Pandora hasn’t proved it can make money, and a lot of folks in radio don’t believe it can and are in fact betting it can’t. In the most recent quarter, Pandora lost $6.8 million.

In theory, anywhere you can aggregate an audience the size of Pandora’s you should be able to monetize it; the more listeners, the more attractive your service is to advertisers. By that logic Pandora should be printing money.

Except that the cost model for streaming music is exactly the opposite of broadcasting. The costs for terrestrial radio stations are essentially fixed; a license, transmitter, studio and staff. Those costs are the same regardless of whether they have one listener or 1 million.

But when it comes to streaming, each new listener means additional costs for bandwidth and music licensing.

And while bandwidth costs are regularly shrinking, music licensing, which accounts for nearly 50 percent of Pandora’s current expenses, is set to increase significantly over the next five years.

According to the terms of the current agreement between webcasters and SoundExchange, which collects and distributes royalties to copyright owners of sound recordings, Pandora currently pays $0.00102 per song it serves through its free streaming service.  By 2015 that figure will increase to $0.00140.

Songs served on Pandora’s commercial-free premium service currently cost the company $0.0017 per play. That will grow to $0.0025 by 2015.

Pandora claims to account for half of online radio listening. Having a big online audience is expensive, and the bigger it gets the more it costs.

That’s a concept that’s hard for broadcasters to get through their minds, and it’s not one they’re willing to try until someone shows it can be done. They’re certainly disinclined to try it during these times, when radio has been struggling because of the protracted ad recession.

So in the end it all comes down to how and whether Pandora pulls it off.

If Pandora fails, all those radio folks who bet against it will look real smart. If Pandora succeeds, making its unique business model work, the radio industry may be in for a gigantic shakeup.