His point (if he'll allow me to paraphrase) was that if advertising works -- i.e., if it brings in more money than it costs -- why not spend everything you have on advertising? If you can trade one dollar for two, why not do it with all your dollars? The way Seth put it was this:
So, why, precisely, do you have an ad budget?
If your ads work, if you can measure them and they return more profit than they cost, why not keep buying them until they stop working?
It's a great point. And I think I have an answer.
Advertising works in two very different ways -- short-term and long-term. Short-term, advertising creates sales. Long-term, advertising is insurance.
The sales aspect is what Seth is referring to. It's a short-term job with a reasonably straightforward ROI. It either returns more than it costs or it doesn't. It's not always easy to measure this ROI because business is so complicated these days that unless you're a direct marketer it's hard to isolate any one factor. Nonetheless, most companies can get a general sense of whether the advertising they are buying is returning more than it's costing.
Coca-Cola doesn't cost 50 cents a can more than Safeway Cola because of any ad they're running today. It's primarily because of all those ads they ran all those years ago.
How Advertising Works: Part 5 - The Final Principle