Making data analytics work: Building a data-driven organization
Summary
TLDRThe transcript emphasizes the importance of organizational design, people, and processes alongside traditional technology hardware and software for effective big data analytics. It highlights the need for data hygiene, governance, and a service bureau culture to ensure analytics are useful to internal customers. The speaker suggests focusing on creating a service bureau culture before deciding on centralization or decentralization. Key roles such as data scientists, business solution architects, campaign experts, and advanced modelers are crucial for leveraging big data effectively, ensuring models are actionable and results-driven.
Takeaways
- 🛠️ The script emphasizes the importance of traditional capabilities such as technology hardware, software, and applications in building infrastructure for big data and analytics.
- 🏗️ It highlights the need for building capabilities around organizational design, people, and processes to effectively utilize big data and analytics.
- 🧑🔬 The role of analysts and data scientists is crucial, and it's important to determine the type of expertise required, whether it's just analysts or more specialized roles like data scientists or business solution architects.
- 🔍 The script stresses the importance of data quality, stating that 'garbage in, garbage out' still applies, and that data hygiene and cleansing are vital for achieving accurate results from analytics models.
- 🗃️ Data governance is essential to ensure that data is clean, accurate, and secure, with clear responsibility assigned for maintaining this standard.
- 🤝 The concept of a 'service bureau' culture within the analytics organization is introduced, where the focus is on providing useful and utilized analytics to internal customers.
- 🌐 The script discusses the balance between centralizing and decentralizing analytics, suggesting that the first question should be about creating a service bureau culture rather than the structure itself.
- 🤔 It suggests that companies should consider the advantages of centralization, such as a culture that rewards and protects advanced data scientists, while ensuring their work is used effectively.
- 💡 The importance of having a framework to help companies decide how much to centralize or decentralize is mentioned, indicating that there is usually a mix of both approaches.
- 📍 The script identifies 'kingpin roles' such as data scientists, business solution architects, campaign experts, and advanced modelers as critical for leveraging big data effectively.
- 🛑 It concludes by emphasizing the need for these roles to ensure that models are not only created but also turned into actionable campaigns that impact consumers or customers.
Q & A
What are the traditional capabilities needed for building big data and analytics infrastructure?
-The traditional capabilities include technology hardware and software, as well as the applications that people typically turn to.
Why is it important to build capabilities around organizational design, people, and processes in big data and analytics?
-It is important because after building the infrastructure, organizations need to determine who will use the data, how it will be used, and what kind of analysts and roles are required for effective data utilization.
What roles are essential in ensuring that big data and analytics are used effectively within an organization?
-Essential roles include data scientists, business solution architects, campaign experts, and advanced modelers, who are responsible for creating models, ensuring data accessibility, and turning models into actionable results.
Why is it crucial to have a service bureau culture within the analytics department?
-A service bureau culture ensures that the analytics team focuses on providing useful and used analytics to internal customers, rather than just doing analytics for the sake of it.
What is the significance of data hygiene and data governance in big data analytics?
-Data hygiene and governance are crucial because they ensure that the data used in analytics is clean, accurate, and secure, which directly impacts the quality of the results produced by the models.
What are some advantages of centralizing analytics within an organization?
-Centralizing analytics can allow for a culture that rewards and protects advanced data scientists, ensuring that their work is used and useful within the organization.
What is the first question an organization should ask itself when considering centralizing or decentralizing analytics?
-The first question should be how to ensure that the organization responsible for analytics views their role as a service bureau and provides useful analytics to internal customers.
What are the key roles that are critical for leveraging big data effectively and quickly?
-The key roles include data scientists, business solution architects, campaign experts, and advanced modelers, who contribute to the creation, accessibility, and application of big data models.
How does the concept of 'garbage in, garbage out' apply to big data and analytics?
-The concept applies as poor quality data input into a model will result in poor quality output, regardless of how sophisticated the model itself is.
What is the role of a business solution architect in big data and analytics?
-A business solution architect is responsible for understanding and creating the right big data warehouse, ensuring that the data is accessible and useful in a usable way.
What is the importance of campaign experts in the context of big data and analytics?
-Campaign experts are crucial as they ensure that advanced models are turned into actionable campaigns that can effectively engage consumers or customers.
Outlines
🤖 Building Organizational Capabilities for Big Data Analytics
The paragraph discusses the multifaceted nature of establishing a robust big data analytics infrastructure. It emphasizes that beyond the traditional focus on technology hardware and software, there's a critical need to build organizational capabilities around design, people, and processes. The speaker highlights the importance of determining who will utilize the analytics, the kind of analysts required (data scientists or just analysts), and the necessity for a business solution architect or a simple database. Furthermore, the paragraph underscores the significance of data quality, data governance, and ensuring data hygiene to avoid the 'garbage in, garbage out' scenario. The speaker suggests that companies should focus on creating a service bureau culture within their analytics teams to provide useful and actionable insights to internal customers. The paragraph also touches on the debate between centralizing and decentralizing analytics, advocating for a balanced approach with a focus on the right service bureau culture.
Mindmap
Keywords
💡Technology Hardware and Software
💡Applications
💡Organizational Design
💡Data Governance
💡Data Hygiene
💡Data Scientists
💡Business Solution Architects
💡Campaign Experts
💡Advanced Modelers
💡Centers of Excellence
💡Service Bureau Culture
Highlights
The importance of building capabilities around organizational design, people, and processes in addition to technology hardware and software.
The necessity of determining who will use big data and analytics and how it will be utilized within an organization.
The role of analysts and data scientists in an organization's big data and analytics strategy.
The need for a business solution architect to understand and create the right big data warehouse.
The significance of data governance in ensuring clean, accurate data for analytics.
The concept of 'garbage in, garbage out' and its relevance to data quality and model outcomes.
The importance of data hygiene and cleansing for effective big data analytics.
The debate between centralizing and decentralizing analytics within an organization.
The suggestion to first focus on creating a service bureau culture for analytics within the organization.
The advantages of centralizing analytics, such as fostering a culture that rewards and protects data scientists.
The critical roles of data scientists, business solution architects, campaign experts, and advanced modelers in leveraging big data.
The role of campaign experts in turning models into actionable campaigns that impact consumers or customers.
The framework for companies to determine the right balance between centralization and decentralization of analytics.
The idea that the first question should be about ensuring the analytics team views their role as providing useful analytics to internal customers.
The emphasis on the importance of the business solution architect in making data accessible and useful.
The discussion on the mix between centralization and decentralization, suggesting there is no one-size-fits-all approach.
Transcripts
the capabilities needed include the
traditional things that people run to
like technology hardware and software
and what kind of applications but the
truth is there's a lot of capabilities
that have to be built around the
organizational design people and
processes because the truth is when you
finish building your infrastructure and
you've got big data and big analytics
you've got to figure out who's going to
use it how's it going to be used
what kind of analysts are you're going
to have do you need data scientists or
just analysts you need a business
solution architect or is it a simple
database in addition to that how do you
make sure that the data you get is good
clean data you know the old garbage in
garbage out still applies garbage in to
a great model the model itself doesn't
give great results so data hygiene data
cleansing making sure data is clean and
having data governance around who is
responsible for keeping and securing
clean accurate data that gets fed into
the big data analytics becomes very very
important so it's more than just the
hardware infrastructure software and
applications it's the people and the
governance around it and we see
companies and clients working towards
centers of excellence and distributed
centers of excellence but they focus on
that kind of breadth a lot of our
clients start there should we centralize
our analytics so should we decentralize
them I don't think that's perhaps the
first question I think the first
question is how do you make sure that
the organization responsible for the
analytics looks at their job as a
Service Bureau and makes sure that they
are providing useful and used analytics
to internal customers there are
advantages to centralizing as an example
some of our more advanced data
scientists you know they're their
definition of a funny joke is about SAS
and sequel most of us don't get that
they want to be in a culture that makes
them rewarded and protected but they
want to see their analytics used and
useful and so yes you can focus on
centralized and decentralized but first
we would suggest focus on are you
creating the right service bureau
culture are your analysts doing
analytics for analytics sake or to help
the business and who in the business and
does that business user believe they
were
helped so that's a framework that we
find is very helpful as companies sort
through how much decentralize and how
much to decentralize and it's always a
mix between both there usually are four
or five kingpin roles where a tremendous
amount of deep expertise can be shared
and in a way that helps a lot of
internal constituents and customers
those roles really fall into the data
scientist category business solution
architects campaign experts and advanced
modelers those roles are really critical
the business solution architect is
someone who's going to really understand
how to create the right big data
warehouse so you can use the data and so
that it's accessible and useful in a
usable way the data scientists our folks
that are going to really help create
that advanced modeling but also they're
going to be able to programmatically
take those models and make sure they're
repeatable and use use programming
language to reduce some of the human
interaction the campaign experts are the
folks that that's that last mile if you
got a great model but you can't turn the
model into a campaign that touches a
consumer or a customer you've got
nothing so campaign experts are that
last mile that make sure that the models
get turned into results that turned into
actions and those are some of the key
roles that are really important to make
sure that a client can leverage big data
effectively and quickly
Browse More Related Video
ML Engineering is Not What You Think - ML jobs Explained
Data Structures & Algorithms Roadmap - What You NEED To Learn
Roles in the data governance domain - organizational roles and data governance roles
How AI is Changing the Professional Sports World
The future of advertising is too fast, too furious | Stelios Anastasiades | TEDxUniversityofNicosia
Cloud Computing Explained
5.0 / 5 (0 votes)