The problem is that explaining the upfront cost to 'the business people' is an issue. In many cases, budgets for dev are yearly and that is kind of an horizon f

The problem is that explaining the upfront cost to 'the business people' is an issue. In many cases, budgets for dev are yearly and that is kind of an horizon for most people (even some who should know better), so even if i'm going to say that your TCO is probably (which is a problem in itself, but probably statistically) going to be lower with formal methods than without, that assumption is spread over many years. There are enough systems that require vast refactoring every year because they were badly done in the first case, but that's the concern of every year's budget/manager, not of the one sitting the current year. That goes for verifying code as well, however that is even harder to sell as it's so much more expensive and even though I 'feel' it is a good idea, I cannot really estimate if the TCO would be lower than just patching with bandaids over the years.