Data Science and Analytics Strategy – An Emergent Design Approach

Much of the discourse around building an organisational data science capability focuses on technical aspects of the discipline. Although many articles mention the importance of gaining management support and managing change, these are often treated as secondary to technical matters. In a soon-to-be-published book, Alex Scriven and I take a contrary view: that building a data science capability is, at its heart, a problem of organisational change.

(Note: A free sample, which includes the front matter and the first chapter of the book, can be accessed here).

The reason many data science strategies tend to be technology-centric because that seems to be the obvious place to start. Consequently, they tend to focus on platforms and processes rather than people and problems. Our experience suggests this is a mistake: building or buying technical capability in itself is no guarantee of a good return on investment. In the book, we describe in detail how to build a people and problem-centric data science capability in a manner that involves minimal upfront investment and improves the chances of tangible returns. Along the way, we provide actionable advice on technology choices and setting up processes ranging from problem formulation to model building and deployment.

The approach described in the book is based on the principles of Emergent Design which we elucidate in the book (for more, see this post on my blog?or ?this paper and this thesis by David Cavallo). The basic idea is to start from the current organisational situation and develop data science capabilities in an incremental way that tackles problems which deliver immediate value. Although the approach has a sound theoretical grounding (which we discuss) our book is not an academic treatise:?we draw on examples from our collective experiences and those of many other data leaders to demonstrate how one can build data science and analytics capabilities using Emergent Design (Note: biographies of the individuals we spoke with are included in the sample).

Since the proposed approach focuses on people and problems rather than platforms and processes, it takes into account factors that are often seen as peripheral to building technical capabilities.?This includes things such as the retention and development of key staff, democratization of data skills, when and how to engage data consultants, and - perhaps most importantly - issues pertaining to AI / data science governance and ethics. The latter are rapidly gaining legislative attention and will therefore be key considerations for leaders and strategists in the years to come.

The approach described in the book has been applied in diverse organisations ranging from multinationals to start-ups and not for profits to government agencies. When speaking with data leaders from different domains, it was heartening to hear that many of them take a similar approach in their own work. Emergent Design is a thing, it seems, so it is somewhat strange that there isn't much written about it. We suspect this is because the details of what one must do depend largely on the context in which one operates. Although this is true, there are broad, context-independent principles or guidelines that can help increase the chances of a successful and sustainable outcome. Our book describes these principles and discusses how they can be applied in different situations.

The book can be pre-ordered on the publisher's website at:

https://www.routledge.com/Data-Science-and-Analytics-Strategy-An-Emergent-Design-Approach/Awati-Scriven/p/book/9781032196329

Thanks for your support!