Maintaining a position is not the same as building a position. If Barksdale is right then the existing players in a market need to work to ensure that they can't lose market share from new entrants.

If a player M controls some "share" in the game and an independent contender C can appear and match M's current share on <1/10 M's current costs, then M's costs to maintain its position relative to all others are not greater than C's costs. If M₀ is M's current offering, then M can carve out part of its budget to fund the entry of their own, controlled Cₘ, independent of M₀, and when Cₘ reaches maturity and matches M's share from M₀ alone, they can let M₀ die—or keep both M₀ and Cₘ around, having advanced their take with greater efficiency than all the costs it would have required to make the same gains with M₀ alone. This assumes that the share from Cₘ comes entirely from those competing with M. This won't be the case; M₀ itself will suffer. But even in the worst case scenario, the share from Cₘ can come entirely from eroding the share of M₀, i.e., a simple transfer. This also won't happen. However, even in the event that it does happen, it's still a net win for M, having used Cₘ to cut their costs by >90% under the regime with M₀.