This is an interesting question. My understanding (from a discussion a year ago with MA) is that the principle question of franchise/branch is one of capital. Branches are more profitable on a store-by-store basis, but rollout is slower because it is capital intensive.

Bringing in advertising to the equation is quite interesting, since it is also capital intensive. At first, I was surprised but upon further reflection it makes sense to me. Subway is pursuing a "weed" strategy, with cheap seeds (franchises) and lots of water and fertilizer (advertising).

As for Starbucks, we found some merit to the "flood the zone" strategy at Oliver, Wyman & Co. If you compare the share of branches to the share of deposits within a market, you can observe an "s" pattern. At first, the marginal deposits increases as you add new branches. At an inflection point, the marginal deposits start to decline per branch and taper off quickly.

With this knowledge, the smart bank can build the right number of branches within a market to capture the optimal share of deposits per branch. For starbucks, this might mean that they might capture %60 of the coffee sales within a particular neighborhood just by owning 40% of its coffee shops.

Adding McDonalds to the equation is even more interesting, since they have a mixed franchise/branch strategy. How do they decide what to do in each market? I don't know.

But I think it points to a flawed assumption in the article: just because the business is similar (or identical in McD's case) doesn't mean that you should necessarily expect to see just one strategy in the marketplace.