I've been talking to quite a few people recently about using Lean Startup (Eric Ries) and the associated Minimum Viable Product philosophy in a variety of product development situations. Whilst being an avid supporter, my concern is that there is another kind of MVP lurking beneath the surface - a Maximum Viable Product.
An evolutionary process that leverages progressive learning makes sense in many situations where new ground is being broken and there is a high risk that a solution might be unobtainable (at least in the form initially envisaged). Clearly, startups do not have a monopoly on this type of development and such an approach has been used for many years in the defence and oil & gas industries (to name but a few).
Whatever the reason for adopting an iterative approach, we must keep sight of the end-game (from all perspectives) and we need to ensure scaling and operational issues are adequately designed for and tested as early as possible. Potentially this might seem contradictory to the idea of a minimum viable product and leads us to the difficult task of probing the envelope of how a maximum viable product might behave.
The catch here is that no amount of customer collaboration will anticipate the issues if everyone is focused on celebrating progress against the original problem statement and not thinking about new operational paradigms that might become necessary to make everything work. The important thing to remember is that product success is dependent upon the customer's ability to realise the benefits using systems that are, most likely, not part of our scope. This is particularly relevant where a single customer might envisage using many instances in concert to solve their underlying problem.
Whilst the Lean Canvas and similar tools can help keep everyone on the same page it is also necessary for us to step off that page and help the customer re-frame their problem space to anticipate. They need to consider where we night be able to go and what that would mean for their supporting systems. Moreover, we need to do it early enough to ensure critical decisions related to fundamental product architecture and direction do not preclude any necessary adaptation.
It can be extremely difficult to respond effectively to a late pivot in the problem space.