[webmaven]

> That’s why I tend to gravitate towards starting migrations with the harder parts of the system. It’s a more challenging project, yes — but you will never have more resources and more support for a modernization effort than you do in the very beginning. The hard bits won’t be any easier months or years later.

It should be no surprise that 5his advice is entirely in the spirit of the MVP. MVPs are supposed to de-risk your startup, and you should therefore always tackle the highest risk part of your plan. If the highest risk part is customer acquisition based on your UVP, buy some advertising and see if potential users click 'buy'. If it is retention, you need an app that they use, or a Wizard of Oz prototype. If it is technical feasibility, you need to build a POC.

For a modernization project, if the riskiest parts are the 'hard bits', then your modernization MVP should absolutely tackle those first.

In practice, though, the riskiest part of a modernization effort is usually whether you have enough political capital to see it through. Or to put it another way, whether there is enough buy-in by management. Which often means that you should first tackle the most recalcitrant parts of the organization that have to sign off on it.