The count of Masters degrees with program titles related to Data Science and Analytics may suggest the maturity of a tech ecosystem…and where large tech companies are more likely to invest.
In compiling a list of post-secondary programs across Canada, we see that both (the provinces of) BC and Ontario have several Masters’ level programs that reference “Data”, “Data Science”, or “Analytics”. By comparison, Alberta currently has no Masters level programs identified as such.
Interestingly, Toronto, Ontario has made it on the shortlist of Amazon’s bid competition for its second head office in North America. And Vancouver, BC just announced that Amazon will generate 3,000 new jobs in that city related to e-commerce, cloud and machine learning.
Ideas expressed in this blog were inspired by Gillian Tett’s* book, The Silo Effect: The Peril of Expertise and the Promise of Breaking Down Barriers. Learnings were added from launching a data analytics firm in the middle of an economic downturn.
silos create barriers in organizations, sectors, and industries
silos interfere with organizational effectiveness
silos can be interpreted as narrow, specialist groups*
silos can also be comprised of systems and data
silos exist in all forms of organizations (profit, non-profit, gov. etc)
experts contribute to creating and reinforcing silos
experts can become vulnerable to job loss in economic downturns
digital data is double-edged, as it both creates and breaks down, silos
data analytics platforms can bridge the gaps across silos
applied data science is an interdisciplinary field that can bridge silos
Silos emerge naturally in all cultures and social settings. The main reason is that we as humans need to classify and group everything around us, but we are not wired to do it the same way. * So we come up with our own interpretation of the world, our work, and our personal lives. Unique interpretations create barriers.
And when we organize people to get things done, like in business, we’re inclined to create functional silos (such as finance, marketing, operations, IT etc.) because it’s efficient to do so. Functions are areas of specialty and are sometimes necessary because of regulation (e.g. professional designations for engineering, law, accounting). But they can be loosely formed too and have no designations per se; as in the case of individuals or teams that simply hoard data for their own benefit.
Whether they emerge organically or by design, the more specialized the silos, the more rigid they become. We see this in functions and disciplines that require deep expertise that takes years to accumulate. The silos lead to barriers when their concepts, jargon, and beliefs become inaccessible to other groups.
The proliferation of silos leads to difficulties among groups trying to work together. People may not understand each others’ roles or ideas; processes may break; opportunities may not be realized. This is problematic in a variety of situations where coordination is essential such as disaster response; product development & launch; construction; healthcare; customer relationship management; energy development etc. Tett* has numerous interesting examples of silos and overcoming their limitations.
Specialists within silos are vulnerable to job loss when the demand for their skills weakens. A good example of this is in the Canadian oil and gas sector. As the economic conditions drop, there is less demand for the associated experts. This chart shows the dip in Alberta-based energy & resource-related jobs as those sectors declined during 2 recessions. There are numerous accounts of how tradespeople and professionals alike lost jobs and have had difficulty transitioning to other sectors.
Alberta, Canada Job Loss In Two Sectors – Indicated by Blue Arrows (see reference at bottom of blog **)
Silos are not limited to technical sectors. The vast array of digital data in all aspects of society propagates silos. Data amplifies who we are as it contains the language and the values that we use to describe the world around us. Imagine your own computer files (pictures, social media, email, documents, spreadsheets etc.) and how they are organized across multiple devices (computer, phone etc.)–are you in control of all that data? Do you know where to find what you need when you need it? Now imagine what it’s like across a team, an organization, a sector, an industry, a jurisdiction etc.
When organizational and data silos grow unabated, they tend to control us through unnecessary complexity. But there are ways to reduce the negative impact and create value. Drawing on Tett’s advice*, here are a number of suggestions:
first and foremost, acknowledge that silos exist and emerge naturally
counter silo-thinking by actively learning beyond your area
make organizations more fluid (such as rotating staff to different areas)
create situations for social collisions such as conferences, co-working spaces
incentivize sharing of information beyond silos
encourage a culture of interpreting information
develop some “cultural translators” who are adept at moving across silos
use computers and data to break down barriers
On the subject of using computers and data, there are at least two key areas to address: Data Analytics Platforms and Data Science (both of which are interpreted broadly in this article).
The use of a Data Analytics Platforms enables data to be integrated from different sources and to be simplified and analyzed–in short, this type of platform helps translate data into information/insight–regardless of the silos from which it came. This idea is not new, but it’s getting easier of late with advancements in cloud, self-serve data analytics, flexible databases, big data technologies, availability of data etc. Those familiar with business intelligence (BI) understand how traditional approaches have been giving way to new capabilities.
Applied Data Science is interpretedas a broad, interdisciplinary field*** that seeks to derive meaning and understanding from digital data. Some of the competencies you might see at play are identified below. Often teams are needed to bring these skills to bear, as it is too difficult to have individuals with such breadth. In this model, the Functional Knowledge pertains to any area being analyzed (e.g. finance, marketing, manufacturing, operations). And the Narrative is the story that pulls it all together, offering an explanation of what the data are saying.
The promise of data analytics platforms and the application of data science is linked their ability to answer questions. To do this responsibly, of course, ethics, policy and the law are essential boundaries for the work. Solving problems and finding opportunities won’t happen by accident; one has to be purposeful and careful to do it right.
As Tett* says, “Data does not reorganize itself, or break down silos by itself; somebody needs to program the computers. What is needed above all is a big dose of human imagination.”*
Applied Data Science is a field that we can look to for those skills. And the interdisciplinary nature of data science may offer new opportunities for specialists (such as engineers, geoscientists, lawyers, accountants etc.) who are dislodged during economic downturns.
*** The use of Applied Data Science is similar but a bit broader than the definition offered here Data Science Definition. “Applied” is used to suggest the practical aspects of what one might see in the workplace.