How to Integrate the Work of the Data Scientist into the Company's Strategy

This blog entry continues the topic of how a Data Scientist can convince colleagues to become more data driven.  The previous blog covered office politics.  This entry covers integration with the strategy and, more specifically, the process that creates the company's strategy.

Every company is unique and consequently so is its strategy process. At first glance, research on how companies develop strategies is complex and contradictory. There are simply too many ways to go about it: SWOT, Five Forces, Contingency Theory, Product Evaluation, etc... The problem here is that choosing one approach means neglecting other approaches and aspects of strategy. In 1999, Henry Mintzberg and Joseph Lampel performed a survey of scholarship in this field. These two researchers categorized their findings into 10 schools, which are:

                1. Design School - sees a fit between environment, customers, and products; uses SWOT Analysis

                2. Planning School - strategy exists in directives and tasks; uses objectives, budgets, and plans

                3. Positioning School - sees an analytical bent to planning; uses game theory and analytics

                4. Entrepreneurial School - sees leadership as critical to success; uses intuition and visioning to set objectives

                5. Cognitive School - sees interpretation of events as input; depends on creativity; one tool of choice is thought maps

                6. Learning School - sees retrospective sense-making as a first step; uses experimentation; depends on a scientific approach

                7. Power School - sees influence within and without the organization as useful; uses bargaining, partnering, and networking; users could be labeled as Politicians

                8. Cultural School - sees strategy as a social process; uses common interests and integration; users are typically people-oriented and value concepts like trust

                9. Environmental School - sees the environment as the greatest constraint;  uses other schools to develop plans; sometimes called Contingency Theory

                10. Configuration School - sees transformations as key to successful strategies; uses concepts like "turn-around" and "dramatic change"; typically used by change-agents


Luckily for us, the researchers went one step further and laid out the ten schools into a ten-step process. This gives us a model on which to lay any given company's process. The ten steps are not performed in a linear manner. Environmental school is performed continuously. Several other schools, specifically Cultural, Cognitive, Learning, and Power, are performed simultaneously.

With this ten-step process, you can ask yourself which ones your company emphasizes, which ones are implemented weakly, and which ones are ignored. Knowing this allows you to determine where your project(s) fall along the 10-step strategy process. And...Knowing what occurred prior to your project and what could have occurred but did not, gives you ideas for enhancing your results for greater impact. Equally important, knowing what could happen after your project ends gives you insight in selling your project to stakeholders.

For example, let's say you determine your particular data mining project fits with the Cognitive School, but you feel that the Entrepreneurial School was skipped. Also, you feel that the Power School is important to your organization. Then your next step should be to use the Learning School, which logically follows as the next step, as a means for gathering support and political strength for your project. Stating the benefits to your projects to senior decision-makers and explicitly stating how to take advantage of your work grows your personal worth to the company.

Good luck.

Recommended Reading:

The original article by Mintzberg and Lampel is worth reading. They actually used the word Safari to describe their work. It provides expanded explanations of the ten schools and plenty of references to the scholarship used in their study.

Mintzberg, H., & Lampel, J. (1999). Reflecting on the Strategy Process. Sloan Management Review, 40(3), 21–30. 

Views: 1157


You need to be a member of Data Science Central to add comments!

Join Data Science Central

Comment by Stephen Penn, DM, PMP on March 13, 2013 at 9:24am

Your question is very timely and puts the topic of the blog into a different context, which I appreciate.  Yes, the ten schools can be applied to Big Data Adoption. In attempting to establish big data capabilities, you're trying to persuade others in your organization to adopt these capabilities.  This is different than my original point of trying to apply data science to an existing strategy.  Persuading others boils down to making a sales pitch. And, every effective sales pitch has five basic parts: 1) identify the problem, 2) name the solution, 3) validate the solution, 4) state why you're the one to implement the solution, and 5) calculate the cost.  Therefore, the answer to your question means combining the ten schools with the five parts of a sales pitch. 

For Big Data, I would say that you either need to pursue enhancing the existing strategy process or adding a whole new step to the existing strategy process. Your decision depends on your given context. The Design School, which is concept oriented, could benefit from Big Data by clarifying existing understandings or challenge existing assumptions.

Let's assume your company uses SWOT analysis, within the Design School, which identifies all (S)trengths, (W)eaknesses, (O)pportunities, and (T)hreats. Big Data creates opportunities to identify misconceptions and assumptions held by organizational members. Anything that adds to the list of items in the SWOT analysis, or corrects the list, is beneficial. For instance, Big Data could spot new markets, thus adding to the list of Opportunities.  Therefore, your sales pitch may be something like...
Part 1 - Existing SWOT models potentially miss shifts in the market.
Part 2 - Big Data can identify emerging patterns that no other means can because combined data (i.e.- Twitter downloads, Census data, and our own data warehouse) is massive.
Part 3 - Other companies are using Big Data (supply whatever specifics you can here, i.e. Hadoop, Julia, etc...) for social network analysis.
Part 4 - I understand how to implement this technology because I understand the data, the technology, and the business challenges ahead.
Part 5 - I will start small with only a few linux machines and a few feeds from Twitter.  Then demonstrate the capabilities in incremental steps.

You may want to read more about the ten schools than the article referenced in the original blog offers. The authors of the article turned their research into a book that goes into much more detail and explain each school in a separate chapter.
Mintzberg, H., Ahlstrand, B., & Lampel, J. (1998). Strategy safari: A guided tour through the wilds of strategic management. New York, New York: Free Press. 

Comment by Joshua Burkhow on March 7, 2013 at 8:35pm

Interesting stuff here Stephen! Any thoughts for Data Scientists trying to get their organization to adopt Big Data initiatives using this strategy approach?

Joshua | DataEnthusiast

© 2021   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service