Role of Leadership in Data Analytics
We started this article series with the title “Data as an Asset.” It is avidly apparent that organizations will be able to grow and thrive only if they integrate data initiatives into their strategic road-map. An organization’s business leadership plays a critical role in the extent of the successes or failures of its data initiatives. The employees of the organization take cues from the business leaders in how they take advantage and extract value from data to drive business outcomes.
Who are the Leaders?
A typical organization’s leadership comprises of:
– The CEO
– P&L leaders of each of the sub-businesses
– Fulfillment leaders for the different process functions like VP-Manufacturing, VP-Sales and VP-R&D/Engineering/CTO.
– Shared Services Leaders like CFO, CIO, CDO and VP-HR
The CIO, and the new CDO role that has come to prominence in the past decade, are traditionally the ones primarily involved in data initiatives. Going forward, depending only on the IT organization to drive data initiatives will not be enough. All top leaders will have to be very engaged for the organization to extract value from data and use it as a competitive advantage.
Leaders – Are you a Player, Owner or Spectator?
As a leader, what role are you playing to enable the organization achieve its outcomes using data? We will use the sports analogy of soccer here.
Player – This is the ideal scenario where all leaders play an active role. They define the vision, integrate the data strategy into the goals set by the vision, and track toward successful execution. More importantly, they not only use data and analytics solutions to keep themselves informed, but also run their operating rhythms and meetings directly in those tools rather than using paper/slides.
Owner – In this case, the leaders are playing a peripheral role. While they support the data initiatives in principle, their involvement is only to the extent of approving funds justified by a financial business case. They delegate the data initiatives – the CEO delegates to the CIO and other business leaders delegate to their respective IT leaders. They infrequently use analytics tools.
Spectator – This is where the leaders are disconnected. Till a decade ago, almost all the leaders were ‘spectators’ in the data space. While some of them are transitioning to become “players” we still see a significant number of leaders stuck being “spectators.”
Role of CEO in Data initiatives
Very often we see that data literacy has an inverse relationship to the seniority level in the organization.
Recently read a post from John Weathington on “Why CEOs must lead big data initiatives.”
The CEO is the primary driver of the culture across the organization. The organization takes cues from how the CEO behaves. It would be ideal if the CEO does lead the data initiatives in the organization. But the reality today is that it is not often the case due to priorities and time constraints.
So, the CEOs should at least be ‘players’ – portray day-to-day behaviors that support data initiatives, e.g. the monthly/quarterly financial review should be done directly in the analytics platform instead of slides/paper and be based off actual data from the source systems. Just this action alone results in a sea change across the organization where the business users themselves drive data quality in the source systems.
CIO and Data ownership
This section touches on the critical aspect of data ownership as it relates to IT leadership.
I recently attended a panel on Data Governance. There is a prevalent thought that was repeated multiple times – “Business owns data” and “Data Governance should be a joint ownership.” While these concepts may possibly work within organizational silos or small organizations, our experience shows that both these concepts are not practical in the real world to implement a sustaining and enterprise-level data governance framework.
What do we mean by the term “ownership”? Using the context of the RACI matrix, ownership means who is “Accountable.” The RACI framework says that, for it to be effective, there should be only one “Accountable” person.
Let me use an example of customer data to explain further the thought process on data ownership.
As we can see from the below image, in an industrial organization like ours, customer data is used in scores of processes across the enterprise. And, these processes are implemented in numerous software applications. And, the only single team that is involved in all these processes as well as all these applications is the IT team.
So, having a team like a Sales/Commercial team “owning” customer data is not effective. The Sales/Commercial team may focus on, at most, 25% of the attributes. It is not the scope or priority of the Sales/Commercial team to support all these processes.
Even within the IT organization, we often hear the IT Leaders use the phrase “IT doesn’t own data”. Imagine the CFO saying that the Finance team does not own financials. Doesn’t this sound ridiculous?
If so, why is it acceptable for the CIO and the IT leadership to say that “IT doesn’t own data” and “IT is not accountable for data quality”?
What is the “I” in IT? How can a CIO be the C.I.O. if he or she doesn’t take active and primary ownership to drive entitlement in extracting value from data?
This is not to say that all data across the enterprise should be owned by IT. But IT can fill the gaps where the business or functional teams are not ready to step up. IT should raise their hands and become the data owners and drive data quality collaboratively while ensuring transparency. This is with the full knowledge that when the business/functional teams become mature enough to own the data, this responsibility is easily transitioned. On the other hand, if a business team is not effective, then the IT team should pull the ownership back.
Chief Information Office – Changing role of IT
The role of IT has changed significantly. Historically, IT was a cost center – basically a service provider. While there have been many changes in the IT industry over the past 2 decades, I wish to highlight the three aspects that are changing the role of IT in organizations:
– Cloud – IaaS – PaaS – SaaS – The Cloud and the different services – Infrastructure/Platform/Software as a Service – have significantly diminished the need and dependency on the IT organization to provide these services. Business teams can directly go to a SaaS provider and implement business process and deploy them globally without any involvement from IT.
– From the above discussion on data ownership, we explained how the IT team, and hence the CIO, is in a unique position to be the common thread, also called as the “Digital Thread,” to play a pivotal role in driving business outcomes through a more active ownership of data.
– As you saw from my earlier post on “What is Data? An Asset-centric view”, Digital Twins help us drive better outcomes for our customers and their assets. And to enable Digital Twins, IT needs to be in the front and center. So, there is rising expectation, and more importantly an opportunity, for IT to build and monetize products.
With this background, the CIO can either sit back and see the IT organization become irrelevant. Or, the CIO can have the foresight to appreciate this transformation in the IT space and have the vision to grab this opportunity to not only evolve the IT organization to be a value-center – driving productivity, but also have the goal of becoming a profit-center by creating products for the end customers.
Continuing with our soccer sports analogy, the CIOs should be the “team captain” for all data driven initiatives. He/she has to drive the organization across the three fronts using his leadership capital to:
– Get the CEO to be a data “player” – push him/her to actively use analytics platforms in their operating rhythms.
– Get his/her peers – the other business leaders (CFO, CMO, etc.) – to be data “players”
– Be a data champion and make no-brainer investments in data initiatives.
Enterprise Data Architecture
Organizations implement large enterprise systems to achieve business outcomes. Large organizations spend hundreds of millions of dollars implementing and operating enterprise-wide global applications like ERP, CRM, PLM & Financial Management. Besides investments, there are three other inputs that should go into implementing these enterprise solutions:
– Enterprise Business Processes – Definition of end-to-end business processes across the Marketing, Commercial, Design/Engineering, Supply Chain, Services and Financials functions.
– Enterprise Technology Architecture – The IT platforms and technologies that will be used to implement these solutions and, more importantly, integrate the different solutions.
– Enterprise Data Architecture – Ensuring that the right data sets as well as the correct linkages are used in modeling the data across the enterprise – Master Data Management plays a key role here.
Almost all the time, organizations invariably miss on these inputs with varying degrees:
Siloed business process – each function defines business processes that work for their own domain e.g. the Commercial defines efficient processes for their Sales team without regard to how their demand signals will integrate with the Supply Chain processes.
Inconsistent technology architecture – Each implementation team makes independent decisions on their own technology stack. This makes it difficult to integrate at an enterprise level. There are also negative implications to the business community. They have to learn multiple tools for the same activity e.g. for reporting, a business user who spans multiple enterprise solutions will have to learn tool A used for reporting in the ERP, learn tool B used for reporting in CRM and learn tool C used for reporting in Financial Management.
Missing Enterprise Data Architecture – Almost 100% of the time, Enterprise Data Architecture is missed in building these enterprise solutions. e.g. each application team defines their own master data set like Item master, they do not use common reference data like Unit of Measures, Currency, Country/Regions, etc. and also designs applications driven by siloed business processes.
The end result is that the benefits across the enterprise is severely limited. This totally defeats the purpose of implementing the large enterprise system in the first place.
This is good segue into the role of the Chief Data Officer as the “Data Architect” for the enterprise.
Chief Data Officer & Enterprise Data Architecture
The CDO team collects all the data from the enterprise to support the analytics needs of the organization. This team is in a unique position with an end-to-end view of the enterprise data. And this team is expected to build the linkages as part of the “Digital Thread” and clearly understands the challenges and difficulties when the different enterprise source applications are not architected properly from a data standpoint.
The CDO organization should be positioned as the Enterprise Data Architect. This team should be engaged in the design and build phases of the enterprise source applications. They should ensure that the source application data model has as all the necessary linking parameters as well as uses the right data objects including Master and Reference data.
Other Leaders in the Organization
Last, but not the least, we need to drive the engagement of the other leaders in the organization including the P&L leaders of each of the sub-businesses, Fulfillment leaders for the different process functions like VP-Manufacturing, VP-Sales and VP-R&D/Engineering/CTO and Shared Services Leaders like CFO and VP-HR. We would like all of them to be «players» in the data initiatives.
Historically all these leaders were «spectators». Recently some have made the painful transition mid-way to being reluctant “owners.” They have started to support the data initiatives. However, they are doing so due to peer pressure, either because competitors or others in their respective industries are doing so, or because of the media hype on Big Data and Machine Learning, which seems to be the recent buzz word or flavor of the day. They lack belief and conviction and fail to appreciate how data can really help them grow their business.
The CIO and the CDO should engage with these leaders to push them to be active «players». Their active engagement is critical for the enterprise-wide success of the data initiatives in delivering outcomes for the organization.
Disclaimer: The opinions expressed in this article are the author’s own and should not be attributed to the author’s employer.
Fuente: Santosh Kudva
9 Distance Measures in Data Science
1. Euclidean Distance
We start with the most common distance measure, namely Euclidean distance. It is a distance
6 Ways AI is Transforming the Finance Industry
Scope of Artificial Intelligence in managing Finances
The sector has been witnessing unprecedented growth in term
Can You Trust Your Data?
The Data Trust Gap
In The State of Data Management — The Impact of Data Distrust, a rec