Last year, I briefly touched upon this concept as an approach to cost effective programme design that was still flexible enough to provide room for iteration for best fit.
Today, I want to explore the concept further to evaluate its potential as a framework for incorporating the concurrent shift in development thinking towards Value for Money (DFID) principles, in addition to designing for best fit.
Value for Money as a Process Driver
Value for Money (VfM) is not the same as traditional monitoring and evaluation which seeks to measure impact of a project, and occurs usually after the fact. In many large scale projects, this may not happen until years after inception.
Instead, VfM is defined by the UK’s National Audit Office as ‘the optimal use of resources to achieve intended outcomes’, which in turn, the DFID document contextualizes for their aid programming investments as “We maximise the impact of each pound spent to improve poor people’s lives.”
If this applies to all investments in aid related programme development, then it follows that it must also apply to earliest stage of discovery and exploration that leads to problem framing i.e. the necessary groundwork to write a comprehensive and inclusive design brief for future programming.
Thus, the conceptual approach that I introduced at the beginning of this post, which is taken from the discipline of Operations Research, and seeks to solve the challenge framed so – what is the optimal solution that minimizes resources (inputs) for maximum outputs (value creation) – fits as a potential framework that can theoretically apply from the earliest stages of implementing development strategy, even before inception of any related projects, including early stage research and feasibility studies. After all, the function of Linear Programming is optimization.
Note: Here I will only consider the theoretical aspects from the point of view of programme design research and development, and not the mathematics. That will have to wait until I have gathered enough data for validation.
Design Research for Programme Design Purposes
In this context, the primary function of such an exploratory project is to identify the opportunity spaces for interventions that would together form an integrated programme designed to effect some sort of positive change in the ecosystem within which it would be implemented, and offer a wider (more inclusive) range of cross-cutting benefits.
In the language of product development, we are attempting to build a working prototype. We cannot build and test first prototypes to see if they work, directly, because our room for failure is much less spacious for experimenting with aid related programming, ethically speaking. This is not a laboratory environment but the real world with enough challenges and adversity already existent.
Programmes are not the same as consumer products, nor are they meant to be designed and tested in isolation before being launched for pilot testing in the market. Their very nature is such that innocent people are involved from the start, often with a history of skepticism regarding any number of well meant donor funded projects aimed at improving their lives. This changes the stringency of the early stage requirements for design planning.
At the same time, the nature of the task is such that no first prototype can be expected to be the final design. So, from the very beginning, what we must do is set the objective of the outcome as a Minimal Working Prototype (MWP) that meets all the criteria for an optimal solution, and NOT a Minimal Viable Product (which may or may not work wholly as intended until tested in the field for iteration.)
That is, the first implementation of the iterative programme design must fall within the bounds of the solution space – that which is represented by the shaded area in the diagram above.
The Optimal Solution is the Iterative Programme Design
Thus, what we must be able to do at the end of the discovery phase of research necessary to write the design brief, is tightly constrain the boundary conditions for the solution space within which the MWP can then be iterated. This minimizes the risk of utter failure, and maximizes the chances of discovering the best fit, and all of this within the definitions of Value for Money and it’s guidelines.
There are numerous ways to set the goals for optimization – one can minimize resources and maximize constraints, or minimize risk and maximize return on resources invested. These will guide our testing of this framework in field conditions to validate the robustness of this theoretical approach.
In this way, we can constrain our efforts to discover best fit within predefined limits of tolerance, while retaining the flexibility to adapt to changing real world circumstances and progressive transformation of operating conditions.
Best fit, then, becomes less a matter of experimentation without boundary conditions and more a discovery of which of the many right answers – if we take the entire shaded area as containing “right answers” to the problem at hand – help us meet the goals of intervention in the complex adaptive system in an optimal manner.
The point to note from this conceptual framework is that there is never any ONE right answer, so much as the answer will be that which we discover to the question “What is needed right now for us to meet our goals, given these changes since we last looked at the system?”
It is this aspect that loads the burden of a successful outcome on the front end of the entire research and development process, given that framing the problem correctly at the outset is what drives the research planning and steers the discovery process in the direction of relevant criteria, conditions, constraints, and user needs that will not only form the bounds of our solution space, but also act as waymarkers for monitoring change and evaluating its progression.