The following article is an excerpt from the new book Minding the Machines: Building and Leading Data Science and Analytics Teams by AI and analytics strategy expert Jeremy Adamson, published by John Wiley & Sons, Inc.
Organize, plan, and build an exceptional data analytics team within your organization
In this new title, Adamson delivers an accessible and insightful roadmap to structuring and leading a successful analytics team. The book explores the tasks, strategies, methods, and frameworks necessary for an organization beginning their first foray into the analytics space or one that is rebooting its team for the umpteenth time in search of success.
In this book, you’ll discover:
- A focus on the three pillars of strategy, process, and people and their role in the iterative and ongoing effort of building an analytics team
- Repeated emphasis on three guiding principles followed by successful analytics teams: start early, go slow, and fully commit
- The importance of creating clear goals and objectives when creating a new analytics unit in an organization
Perfect for executives, managers, team leads, and other business leaders tasked with structuring and leading a successful analytics team, Minding the Machines is also an indispensable resource for data scientists and analysts who seek to better understand how their individual efforts fit into their team’s overall results. EXCERPT FOLLOWS:
Top Success Factors for Data Scientists
Technical abilities, and especially an intrinsic technical intuition, are table stakes for an analytics practitioner. The work of an analytics team is done when the fingers are on the keyboard, and each team member from the most junior analyst to the leader of the function need to be conversant in the practice, the tools, the languages and ultimately to the application of data science theory to business problems. This theoretical and technical side needs to be balanced with an instinctive drive toward value creation, however, and not simply as technology for technology’s sake. Along the other axis of development, team members need to be balanced between individuals striving for higher levels of management responsibility and those who find deep satisfaction in the work itself. Choosing individuals who are passionate about the field and their personal development while still committed to remaining in the role is important for the stability of the team and the successful delivery of longer-term projects. Success, generally, is a matter of an appropriate balance between these poles of technology versus business and high drivers versus core contributors.
The characteristics of a successful data scientist can vary significantly between organizations. Consultative, broadly educated, and mature practitioners can be successful in legacy environments such as financial services and the public sector. Energetic coders with deep knowledge of particular fields of practice can be equally successful in a startup or technical organization. General success in the field, however, tends to be for those who are somewhere in the middle of the two extremes, adjusting for specific corporate culture and industry norms.
While it is difficult (and often counterproductive) to attempt to distill a person into a series of adjectives and attributes, there are several other features that are important for a successful team. It should be noted that no one individual will have all of these characteristics. Teams are a blend of individuals, often with different traits, which results in frictions that spawn creativity and progress. There are eight soft traits for analytics practitioners that are critical to success:
Curious
Every strong data scientist and analytics professional will, without exception, exhibit a strong natural curiosity about their work and the world around them. They will become animated at discussions of exogenous data sets and speculate enthusiastically about the drivers of consumer behavior. They will notice things that others do not, and they will see connections that others do not. In most cases, it is this deep and instinctive curiosity that has driven them toward a career in data science. Data scientists without curiosity are like ageusiac chefs or dyscalculic actuaries—technically capable, but grossly disadvantaged.
Adaptable
Every strong practitioner will also, again without exception, be exceedingly adaptable in how they approach problems. They will find ways to make things work regardless of obstacles and find great personal satisfaction in doing so. They will not wait for detailed instructions or condemn a project as impossible. They will be stimulated by challenge and happily adjust their approach to deliver.
Consultative
They can build relationships with key people and naturally see that as part of their role. Their previous successes will leverage relationships as much as technology. They will enjoy teamwork and collaborative projects.
Skeptical
They do not take things at their face value and will instinctively distrust heuristics and traditional ways of approaching problems. They will try to disprove assumptions and trust data over instincts. At the same time, they will adjust their own positions without reservation in the face of compelling evidence.
Creative
They will prefer open-ended problems around which they can craft their own solution. They will be personally gratified by original problems and often will have creative personal hobbies or side projects. They are not motivated by money but rather by their passion.
Business Focused
They tend to be motivated toward solving applied business problems. They do not need to fully know the business, but they need to develop an interest in the problems of the business and learn to speak the language of the business.
Egalitarian
They will see analytics as a team endeavor and work with rather than for their direct supervisor. They never wield title as a proxy for authority, nor do they view hierarchy as an indicator of intelligence. They seek the best solution, and artfully challenge people to defend their approach and premises regardless of their position.
Competitive
Though collegial with their in-group and personally caring toward their peers, they will be competitive outwardly. Driven by a need to achieve and deliver value, they will be personally motivated to find ways to execute on projects regardless of issues with dependency or formal authority.
Of these characteristics, the primary are curiosity and adaptability. Without these, all others will fail to foster success, and all others will naturally emerge if these two are present.
There is a deep divide between the business and the analytics function due to a lack of mutual intelligibility. Problems and projects are misinterpreted, and often groups have divergent interests and priorities, which leads to non-operationalizable deliverables. For example, the marketing department may request a segmentation model, the unspoken subtext being that they ultimately wish to understand their customer. The analytics function, not understanding the subtext, responds by providing a table of results that maximize behavioral differences through agglomerative clustering, which is technically appropriate but wholly insufficient for the underlying need of the stakeholder. The absence of a common knowledge base can quickly derail a project. These fundamental differences can balloon into conflict without mutual understanding and without the presence of translators who can serve as intermediaries between these two camps. Certainly, well-rounded teams need specialists with deep knowledge in relevant areas, but blended personalities should be the norm. More-established geeks can often decry this perspective, pointing to celebrated eccentrics in the field, and while it is true that every Jobs needs a Wozniak, it is equally true that every Wozniak needs a Jobs.
The archetypal surly guru, their bitterness tolerated because of their indispensable skills, is gone. In the past this was considered a sign of credibility—that the truly technically gifted needed to be a little bit eccentric, and it was the cost of doing business with these enigmatic sages. Modern data and analytics practitioners need to be professionals and to enable the business through a consultative and mature approach rather than the raw application of technology. The gurus have traded their hoodies for oxfords.
Sign up for the free insideAI News newsletter.
Join us on Twitter: @InsideBigData1 – https://twitter.com
Speak Your Mind