Executives talk about data and want their companies to become data driven. But a lot of companies fail, which is due to various reasons. Becoming data driven is a cultural transformation in a company, not a technical implementation. In this talk, we will look at potential risks and how to implement a data culture.
2. Who is talking to you?
+ Vice President for Data & Insights @ Magenta Telekom
starting in May. Former Head of Data at Uniqa Insurance
Group and A1, Microsoftie, …
+ Book Author (2009: Cloud; 2019: Data Science in the
enterprise, 2022: The Data Science Handbook)
+ Teacher at the Executive MBA “Data Science” at WU
Wien: Data Strategy
+ Speaker at global events such as London Tech Week,
GITEX Dubai, WeAreDevelopers, DSC, …
Linkedin: Mario Meir-Huber
Twitter: mario_mh
Blog: cloudvane.net
4. Reasons
for
failure
Data is not the classic IT:
Data is generated and used
decentrally in the specialist
departments. IT often does
not understand the
complexity behind it
In the specialist departments,
there is often little ownership
of technical data
management systems, which
results in silos
Data quality and a good
architecture do not bring any
measurable added value and
are therefore usually only
approached with limited
budgets. However, if this is
not done, all data initiatives
will fail
Focus on 3 levels
Strong decentralization requires
a culture change, but also
central steering
Level 1 - Technology: having
the Data Platform up-to-date
Level 2 - Culture: Collaboration
in a decentralized setup; new
way of working with data
Level 3 – Governance: Data
Governance & Organisational
Governance
Many companies find it
very difficult to establish
proper data practices
Large consulting
companies promise
extensive financial benefits
through data-driven use
cases
However, their
implementation often fails
Solution
Data
management
issues
in
business
Technology
Governance
Culture
11. A Use-Case driven
approach to the Data
Strategy
Improoving the maturity in Data, Use-Case driven
Based on research at the WU Wien
12. Maturity
Impact High
Low
High
Low
Low hanging fruits
Challengers
Playgrounds
Underworld
Impact:
• Impact can be measured either by
financial impact in $ or strategic
importance. Scale is oriented on the
most impactful project and from 1 to
10
Complexity:
• Complexity is a measure from 1 to 10
in the dimension of Architecture,
Governance and available Skills
Use-Case Repository
13. Impact High
Low
Low hanging fruits
Challengers
Playgrounds
Underworld
Low hanging fruits:
• Ideal projects to execute: low
complexity and high impact
• Data is available and the projects can
be started easily. Typically, these
projects are often seen when the
company already has a high maturity
in data
Use-Case Repository
Maturity
High
Low
14. Impact High
Low
Low hanging fruits
Challengers
Playgrounds
Underworld
Challengers:
• Projects bring great business impact,
but they are difficult to execute (e.g.
Data isn’t available, skills aren’t good
enough, …)
• Before executing these projects, try to
remove complexity!
Use-Case Repository
Maturity
High
Low
15. Impact High
Low
Low hanging fruits
Challengers
Playgrounds
Underworld
Playgrounds:
• Projects that have a low complexity
and can be done easily. However, they
don’t bring much business value
• Avoid doing them, unless for training
purposes or the costs are much below
the impact to achieve
Use-Case Repository
Maturity
High
Low
16. Impact High
Low
Low hanging fruits
Challengers
Playgrounds
Underworld
Underworld:
• Projects have limited business value
and are very complex. STAY AWAY
Use-Case Repository
Maturity
High
Low
17. Impact High
Low
Plotting the Use-Cases to the Quadrants:
• Each Use-Case gets plotted based on
the different measurements
• Use-Cases with the lowest complexity
and best Business Impact get
executed first
• In parallel, it is essential to lower
complexity and move more use-cases
to the low hanging fruits
Use Case A
Use Case B
Use Case C
Use Case D
Use Case E
Use Case G
Use Case H
Use Case F
Use Case I
Use Case J
Use Case K
Use Case L
Use Case N
Use Case M
Use Case O
Use Case P
Use-Case Repository
Maturity
High
Low
18. Use-Case Repository
Impact High
Low
When removing complexity, more use-
cases can be executed:
• Enable the organisation to become
more capable (skill development)
• Raise awareness for new (and
effective) tools to deliver more with
lower effort
• Improve the technical platforms
• Apply Governance that doesn‘t limit
but increases time to execution
Use Case A
Use Case B
Use Case CUse Case D
Use Case E
Use Case
G
Use Case H
Use Case F
Use Case I
Use Case J
Use Case K
Use Case L
Use Case N
Use Case M
Use Case O
Use Case P
Maturity
High
Low
20. Remove complexity by increasing the
maturity in 3 areas
Technology
Governance
Culture
+ Technology: ensuring state-
of-the art technical
platforms
+ Governance: a proper data
governance in a great
organisational governance
+ Culture: changing the
corporate culture to
become data driven
21. Measuring the complexity by Critical
Success Factors (CSP’s)
+ Complexity is ever evolving: what was “state of the art”
might be complex the years thereafter
-> If you stand still, you will actually move “backwards” in technology
+ Literature knows several critical success factors, which are
grouped into the 3 domains
-> Improving with all of them is the key to success
22. CSFs: Technology
+ Technology Infrastructure: what is the status of the technology infrastructure?
-> Usage of Cloud Technology vs. On-Premise Stack
+ Data Models: how is your data modeled?
-> Data Model design, Storage techniques
+ Reporting and Data Science technology
-> What tools are available?
+ Stack integration
-> Is the technology stack integrated into the overall IT architecture?
+ Scalability
-> Can the stack be scaled individually?
+ Service oriented architecture and mindset
-> Data Mesh vs. Monolithic approach
23. CSFs: Culture
+ Skills: How are skills managed within the organisation? Is there an upskilling
program in place?
+ Stakeholder integration: how are stakeholders managed by the data units?
+ Manager’s know-how and support: Do managers have technical and data
understanding? Do they use data for their daily decisions?
+ PMO Organisation: how are data projects managed?
+ Agility: how does the organisation react to change?
+ Communication: How do Business, IT and Data Units communicate?
24. CSFs: Governance
+ MDM: How is Master Data tracked, is it comprehensive?
+ Data Quality: what is the level of Data Quality in your organisation?
+ Data Sharing: How is data shared within your organisation? Is it sharable?
+ Privacy and Security: What is the level of privacy and security in your
organisation for data?
+ Accessability and Searchability: Can you easily search for data? How
accessible is data?
+ Data Ownership: How is data ownership in your organisation? Is there a
decentralized ownership / stewardship in your organisation?
26. Further research
+ We will do further research on the impact of CSFs for
successful data projects
-> not every CSF has an equal impact.
+ Get involved: the framework is getting stronger with more input
from people and organisations
-> we plan to setup a “gremium” to score the impact and then refine
-> we will evaluate the ”level” of each CSF on an annual basis
28. Literature and Further Read
+ Critical Success Factors for Big Data: A Systematic Literature Review (2018); https://ieeexplore.ieee.org/abstract/document/9127414
+ Towards A Process View on Critical Success Factors in Big Data Analytics Projects (2015); https://core.ac.uk/download/pdf/301365683.pdf
+ Determining Critical Success Factors for Big Data Projects (2018); https://www.proquest.com/openview/e92a2045a2dee3fef988de6f294a9f08/1?pq-
origsite=gscholar&cbl=18750
+ Critical success factor categories for big data: A preliminary analysis of the current academic landscape (2017);
https://ieeexplore.ieee.org/abstract/document/8102327
+ Quantitative Comparison of Big Data Analytics and Business Intelligence Project Success Factors (2018); https://link.springer.com/chapter/10.1007/978-3-
030-15154-6_4
+ An evaluation of the critical success factors impacting artificial intelligence implementation (2022);
https://www.sciencedirect.com/science/article/abs/pii/S0268401222000792
+ Big data team process methodologies: A literature review and the identification of key factors for a project's success (2016);
https://ieeexplore.ieee.org/abstract/document/7840936
+ Artificial Intelligence Project Success Factors—Beyond the Ethical Principles (2022); https://link.springer.com/chapter/10.1007/978-3-030-98997-2_4
+ Contextual critical success factors for the implementation of business intelligence & analytics: A qualitative case study (2019);
https://aisel.aisnet.org/cgi/viewcontent.cgi?article=1024&context=confirm2019