Growth through acquisition is a pretty workable model. Especially when you cut the resulting companies to the bone and basically put their product on life support.

For products used by big enterprises its pretty amazing how long a product can go without meaningful updates and still have companies paying for maintenance. I've seen it a few times now, the engineering/testing is basically cut to one developer who's job is basically to fix bugs and nothing else. For many enterprises this is a pretty decent model. In many cases they don't need or want new features, and the IT guys are very happy if they just get a point release once or twice a year that is a 100% drop in replacement with some bugs fixed.

For a large conglomerate that can sell new licenses as part of a bundle deal, its a good plan to, the customer base grows at about the same rate as people drop the products. The result is just a constant cash flow with basically no overhead.