So here's the math I'm doing: switching to GM from the current model -- we'll call it SPP for Sonambulistic Promotional Pitches -- lowers the cost of hiring Saatchi&BBDDBO to write said SPP, and it lowers the cost of the media buy, because who needs a :30 spot on E.R. when you've got micromarkets, right?
However, it replaces these SPP costs with employee time. (Lets call phone fax etc bills a sunk cost.) Now, there are obviously situations where this is a big win -- Southwest Airlines has discovered that by letting its people talk to the customers in their own voice, they improve their bottom line. This is A Good Thing. Southwest has some pecularities to it, however -- its front line people are already in contact with the public, so there's no shelling out extra time; an airline ticket is a considered purchase; and Southwest tickets are cheaper. In otherwords, SWA is not trivially comparable to a) businesses that currently have low public contact (e.g. Intel), b) rely on smaller or less considered purchases (e.g. Staples) or c) have similar or higher prices than their cometitors (e.g. Coke).
In these situations, the resources companies would have to shift out of SPP and into GM would include a considerable amount of employee time, and since every 1000th hour of such time is a half-time job, it triggers the usual welter of W-2 related expenses. Furthermore, since you would want those employees to regard their employer favorably, you would have to pay them more than minimum wage. So lets say you have 1 FTE to devote to your GM strategy, at a salary + expenses of 30K (so this person is making maybe 23K a year -- an entry-level hire, albeit a motivated one.)
E.R. is around a $25 CPM, so for 30K I can reach an audience of something like a million two. Now lets say that a national spot is a blunt weapon, so maybe only 50% of those people would ever even consider buying Jell-o® brand flavored gelatin, and that they need to see my ad 3 times to have it register. This means I'm effectively reaching, for $30K, 200,000 people, some percentage of which will salivate over my ad showing a Regular Mom making Jell-o® brand flavored gelatin Freez-Treets® in the shape of Pol Pot.
Now how many people can our eager $23K a year hire reach? Lets say the average micromarket is 100 people. (I suspect its smaller, but we'll put our thumb on the scale...) and our new hire, Felicity J. Gonzo, has 2K hours/yr to spend with these excited, loyal, and passionately engaged MMMs (micromarket members). Lets say that Felicity has such a winning way with the MMMs that to kno-o-ow her is to lo-o-ove her, so she is 100% more effective than SPP. Ergo, she only needs to reach half as many people to be as effective as my three SPP :30 spots on ER.
To reach 100,000 people 100 people at a time in 2000 hours (assume she never talks on the phone with her friends or uses the company computer to book plane tickets while the boss isn't looking), she has 2 hours per year to spend with each micromarket.
Now you can slice this math any way you like -- pay Felicity less, make micromarkets bigger, give SPP a 10% catchment or 1% even, make Felicity 5 times as effective instead of just double, but there are serious counter-forces as well -- I can run my ad on cable for single digit CPMs, and, most worryingly, I might want to reach 100 million people, not just 1.
This last bit is a corker. The single, critical advantage of the SPP strategy is that, no matter how ineffective each ad is, it scales like a motherfucker. Strategies with low fixed costs and high variable costs always look terrifically competitive when they first launch, but the costs grow like weeds because unlike media, human costs only go up. So my biggest doubt about the GM idea is that, no matter how you arrange the costs, at some point in audience size, probably in the millions and almost certainly in the 10s of milliosn, the money you saved in not paying for your SPP up front blows up in your face in labor costs.