Eric Ries' presentation on "The Lean Startup: a Disciplined Approach to Imagining, Designing, and Building New Products" has three processes depicted for "product" development.? I put product in quotes because I think you can accurately substitute "service" or "process" and the presentation still applies very well to the situation if you assume that some technology or other development activity is required to support your service or process.
Here's the presentation, but allow me to call your attention in particular to slides 20, 21, and 22.? In slide 20, you see the traditional waterfall strategy, which Eric explains is appropriate when you know the problem well, and the solution is also well known and understood.
In slide 21, we see an Agile approach, where the problem is known, but the solution is not well-known and understood.? In fact, a decent number of agile projects get into trouble because the problem is not well-understood and they get caught in what feels like "churn".
In slide 22, we see the introduction of the Customer Development Process to the mix, and the situation is one where the problem is not known, and the solution is also unknown.? So part of the journey is to discover the right problem to solve, as well as the solution that solves it.
I think a lot of BPM projects would do well to adopt this third method (or an adaptation of it) as well.? At a high level we may know that there is a problem (opportunity), but we may not understand what that problem is at a detailed level that would allow us to design a good solution.? So we have to do some level of customer development (in our case these may be users) to understand the problem adequately, and then use an iterative approach to get the proposed solutions improved and adapted as quickly as possible.? At the very least, when taking on a BPM project, one has to keep an open-mind to improvements that haven't been thought of, and to new problem-identification that may happen over the course of the project.
A blurb about the slideshow is available on O'Reilly here.
Also, a more in-depth presentation was done previously on the Customer Development Methodology:
This version goes into a lot more detail on how to build the startup around this methodology, and makes a stronger case for why you need to iterate between business plan and hypotheses before funding, rather than proposing a business plan, getting funding, and *then* testing the market... its a good read!