Phillip:"... realistically — which publications are equipped to say that “Print marketing promo A, resulted in B visits to our Web site, that turned into C subscriptions, D product buyers and resulted in E online ad revenue?”"

From the advertiser's perspective though, if they are smart in their marketing efforts, they should be able to proxy the effectiveness of their ad channels. I had an interesting conversation with a retail veteran once about effectiveness of mass channels (like newspaper ad) & direct channels (like email blast). He said he does not really care about any metrics related to the channels specifically; he just divide the sales he got after the marketing effort by the cost he spent on those channels, and he can conclude that mass is more effective for him.

I like both POV from Adam & Jon. Profitability is how most for-profit define success, so any efforts / costs spent in any marketing channel should directly or indirectly lead to incremental revenue that hopefully at least cover the costs. However, if there are some other definition of success like spreading information to people, measures like followers or friends can be at least a leading indicator if not the absolute measure.

More importantly, new channels sometimes redefine what success can be. In Walrus' case, should being a "magazine" even be part of their mission? Or is it really just publishing quality general interest materials to their target audience through any channel, any format? These all shape what metrics are relevant to their success measures.