I suspect online advertising is undervalued precisely because it can be measured. I further suspect that "brand advertising" is a scam for all but the largest advertisers.
The largest advertisers became the largest because they spend a lot on brand advertising. So perhaps it's true that you have to spend a lot to make brand advertising effective, but it's not a "scam" in any conventional sense of the word. It is what it is.
If we're going to sling arrows at online advertising, I think we have to start with the fact that there are a lot more publishers than there were back in the day, and virtually all of them have an undifferentiated audience of "eyeballs". It's still possible to get high CPM rates, as long as you've got a targeted audience and a sales team.
That (sort of) substantiates the parent post's claim. If only 50% of your advertising money is effective, then spending more will still be more effective, no?
You're making an assumption of corporate stupidity that I find to be highly unlikely. Large companies do not become large companies by tossing hundreds of millions of dollars down dark wells without analysis.
I'm making the assumption that Coke and Nike, et al. are smart enough to have at least rudimentary knowledge of the effectiveness of their brand advertising. Is this an assumption? Yes. Is it a more reasonable assumption, a priori, than the alternative? I think so.
