Analytics and Data Visualization...PROS: Data storage of flat files and Tableau files is possible on Tableau Server. CONS: There is limited functionality and security options here.

Analytics and Data VisualizationWhat to Consider When Scaling up your Team

Table of Contents

01 Analytics and data visualization: the basics

02 How data visualization impacts business growth

03 Data science vs data visualization

04 Tools and software for data analytics

05 Successful implementation and adoption

06 Data analytics training methods

07 Certifications

08 Making smarter, faster, data-driven decisions

09 The future of data visualization and data analytics

Analytics and data visualization: the basics

01

1

Data visualizations—bar charts, line charts, scatter plots—are not just meant to look pretty. Good data

visualizations transform raw numbers into actionable insights by drawing on our innate ability to recognize

patterns and trends. Humans excel at recognizing differences in length, size, and color; data visualizations

leverage these skills to communicate the relationship between values graphically.Your company likely has

millions or billions of data points—enough that without a visual abstraction it would be impossible to gain any

meaningful information from it. A simple bar chart can unlock hidden patterns in your data that numbers on a

spreadsheet would never reveal.

Today, more and more companies are adopting data visualization technologies such as Tableau and Power BI.

These tools allow users to combine large and disparate data sources seamlessly, creating powerful visuals in

the process. By connecting to databases, spreadsheets, and spatial files, companies can create organization-

spanning reports that lead directly to higher sales, less waste, and greater efficiency.

By connecting to databases, spreadsheets, and spatial files, companies can create organization-spanning reports that lead directly to higher sales, less waste, and greater efficiency.

How data visualization impacts business

02

3

The rapid growth of data has been the defining business opportunity for organizations for the last decade.

While data has grown vast throughout organizations large and small the challenge has been the same for all

organizations: how can they scale insights and action from this data.

An early solution to this problem was to invest in data scientists: high-skill workers with a deep understanding

of programming, machine learning, and big data. These individuals remain highly sought after and can solve

extremely complicated challenges grounded in the data, but they do not typically have subject matter expertise

nor can steer organizational change. In order to solve this challenge, organizations must not only hire experts

in data. Rather, they must equip their business teams, including technology professionals, leaders, and their

audiences, with the tools, processes, and culture to effectively tap into their data.

To build a culture that embraces data and scales insights,

organizations must embrace curiosity, lean approaches, and teamwork.

This can be done by adopting self-service analytics tools that make

data exploration straightforward.

This also means transitioning from traditional reporting practices to

agile, iterative approaches. Reports and dashboards can be treated as

iterative, living products derived from data. As part of the process, this

requires using agile approaches and interdisciplinary teams to scale

development.

In order to scale insights and action from data, organizations must

invest in making analytics a central part of their culture. In fact,

McKinsey&Co found that organizations that invest in making analytics

central to their operations will scale their analytics four times faster.

Organizations must equip their business teams, including technology professionals, leaders, and their audiences, with the tools, processes, and culture to effectively tap into their data.

Data science vs data visualization

03

5

What Is Data Science?Data science is the process of collecting and analyzing data in order to get useful knowledge and insight out of it.

Data scientists are experts in statistics, machine learning, and fundamentally solve problems with programmatic

solutions. They work with large data sets and often solve problems with large interdisciplinary teams. Frequently

they work on projects deemed high-value and often in a research and development role.

How Data Science Is Used

Businesses use data science in many different ways to provide better service to their customers. Search engines

use data science algorithms to deliver relevant search results. Streaming services use data science to recommend

television shows they think you’ll enjoy. Even social media sites use data science through face recognition

algorithms to allow you to better tag your friends in pictures. Regardless of industry, data science is fundamentally

anchored in automation. Whether it’s models, predictions, or even summary data, data scientists write code that

can be run at any time so that any process is repeatable at a click of a button. But automation shouldn’t just be left

to a small subset of an organization like a data scientist or data engineer. Automation should be accomplished by

any member of an organization.

You may have heard about data science and data visualization in the context of digital marketing or big data, but what exactly is the difference between these two concepts?

6

Differences Between Data Visualization And Data Science

Whether it’s an experienced data scientist or business analyst fresh out of college, there is a need within

any organization to quickly summarize data for insights. To do so means using data preparation and data

visualization tools to represent data in the form of a chart or graph. When visual elements are used it

allows people to better spot and understand patterns in the data compared to data in a table. Although

data visualization and data science are linked, they are not the same. Data scienceinvolves the entire cycle

of capturing, maintaining, processing, analyzing, and communicating data, and for a data scientist, data

visualization only comes into play in the last two of these stages.

How do Data Scientists Use Data Visualization

Data visualization occurs during the initial part of the analytics stage, allowing

data scientists to represent the data in a way that’s easier to analyze, and

during the communication stage, to inform business leaders in their decision

making. While data science can include insights, explanations, and predictions

about the data, data visualization is a visual representation of trends in

the data itself using charts, graphs, or dashboards. Data science and data

visualization might involve similar elements, but they each bear a distinct role

in the field of big data. Data visualization is a tool within the larger process of

data science, and both are integral in the future of technology.

While data science can include insights, explanations, and predictions about the data, data visualization is a visual representation of trends in the data itself using charts, graphs, or dashboards.

Tools and software for data analytics

04

8

No matter what marketing or sales tell you, there is not one data analytics tool that does everything well. This

can make selecting the right tools (or tools) for your department difficult. Below, we break down common data

software into four grading categories:

Data Visualization & Dashboard Creation Data Blending, Preparation, & Automation

Data Storage / Reporting Data Marts Advanced Analytics

Software OptionsThere are four main tools we use in our consultancy at Tessellation and through our trainings via Data Coach.

These tools are Tableau , Alteryx , Power BI , and Snowflake. Below we’ll compare how each of the tools performs

across four categories:

TABLEAU Connect all of your data with critical external sources like public health status,

POWER B1 Connect to and visualize any data using the unified, scalable platform for self-service and enterprise

ALTERYX Combine data preparation, data blending, and analytics — predictive, statistical and spatial —

SNOWFLAKE Seamlessly share and consume shared data to collaborate across your

9

PROS: Tableau is the hands-down leader in data visualization and has unmatched flexibility. The combination of its features allows for near-limitless possibilities for creating visualizations.

CONS: While it is easy to get started with Tableau, there can be a steep learning curve to become an expert visualization and dashboard creator.

Data Visualization & Dashboard Creation: 10/10

Data Blending, Preparation, & Automation: 4/10

PROS: Within the Tableau Desktop and Tableau Prep tools, you have all of the basics covered: unions, joins, and pivots (just to name a few). It is relatively simple to use this functionality andonly requires a passing understanding of database design.

CONS: There is limited functionality here. Only a dozen or so common operations are supported and data can only be output to flat files or Tableau files, making it impossible to be an enterprise data blend, preparation, and automation tool.

Data Storage / Reporting Data Marts: 2/10

PROS: Data storage of flat files and Tableau files is possible on Tableau Server.

CONS: There is limited functionality and security options here. To be fair, Tableau doesn’t pretend to be a data storage tool. However, it doesn’t stop clients from trying to use it as a reporting data mart. Using Tableau Server as the main location to store reporting data often leads to the proliferation of data and a “wild-west” with limited data governance. It is best to integrate Tableau with a dedicated reporting data mart tool.

Advanced Analytics: 4/10

PROS: There is built-in functionality for integrating some high-level advanced analytics into your visualizations. Cluster analysis, for example, can be added with a simple drag-and drop.

CONS: The advanced analytics and modeling options are limited, and the ones that do exist have few configuration options. There is the option of live integration with other tools, such as Python and R. However, this is rarely recommended for interactive visualizations. Incorporating these integrations almost always results in a slower performing dashboard. Specifically, you can expect to have longer delays for the render times with new queries.

Pricing: 7/10

PROS: Relatively inexpensive per-user pricing, ranging from $12 to $70 per month per user, depending on the role.

CONS: There is a focus on per-user pricing, which means that price breaks are rare for larger organizations. You can expect the price to scale linearly.

10

PROS: Alteryx is the king of batch processing for visualization. Need 1000+ reports quickly, all with the same format but different data elements? Alteryx is the right tool for you.

CONS: The visualization components are tough to use and the output is (generally) unattractive.

Data Visualization & Dashboard Creation: 5/10

Data Blending, Preparation, & Automation: 10/10

PROS: Alteryx is the leader in the data blending, preparation, and automation space. No other tool comes close. With an easy to use drag-on-drop workflow building system, access to hundreds of tools, and a community of thousands of creator-made tools, it’s hard to beat Alteryx in this space. Also, Alteryx can read from and write to dozens of commonly used database systems. You can also customize connections via APIs, ODBC, and JDBC connectors for unsupported systems.

CONS: Though you do not need to be an expert SQL developer, a basic knowledge of database practices and designs is needed to be effective.

Data Storage / Reporting Data Marts: 0/10

PROS: Alteryx doesn’t even attempt to touch this area, which is good. It relies on other systems for storage.

CONS: You will need a landing area for your workflow output.

Advanced Analytics: 9/10

PROS: Alteryx has made a huge investment in abstracting complicated model practices and making it easy to drag-and-drop procedures into a workflow that would usually require an academic data scientist. Absolutely no coding in R or Python is required, although you can if you want to.

CONS: Sometimes Alteryx makes it too easy. Users don’t necessarily understand the background math to know why certain models are appropriate.

Pricing: 3/10

PROS: The pros are generally only a small subset of the analysts at an enterprise actually need access to this tool. Limiting the number of users does help keep costs down.

CONS: Alteryx Designer, their development tool, comes in at a whopping $5.2k per year per user. This is the most common complaint we hear about Alteryx. The Alteryx data add-ons and enterprise tools are more competitively priced.

11

PROS: Power BI makes the creation of visualizations, reports, and dashboards super easy. For visualizations, simply select the type of visual you would like to create and fill in all of the needed information via adding metrics and categorical values to parameters.

CONS: Customization for net-new visualizations that are not “out-of-the-box” can be difficult. It is possible to create or use user-created visualization types, but it is an error-prone and technically difficult process.

Data Visualization & Dashboard Creation: 8/10

Data Blending, Preparation, & Automation: 8/10

PROS: At its heart, Power BI is a data modeling tool. Power BI has a (relatively) simple model building interface where you can set up relationships between disparate sources and then clean and prepare the data.

CONS: Conceptual knowledge of database modeling is a prerequisite for being able to build these models.

Data Storage / Reporting Data Marts: 3/10

PROS: Much the same as Tableau, Power BI is great at storing data in certain file types. Power BI has the added advantage of integrating well with the Microsoft suite.

CONS: There are limitations on file size and data types.

Advanced Analytics: 4/10

PROS: Microsoft is one of the leading advanced analytics firms. In a lot of situations, you can integrate external Microsoft products with Power BI fairly quickly and easily.

CONS: Advanced analytics are not built into the tool itself. You may need another tool, which may be expensive, in order to accomplish your analytics goals.

Pricing: 5/10

PROS: For existing Microsoft users, Power BI is anywhere from cheap to free.

CONS: Power BI can be hard to integrate with non-Microsoft tools. It can also be difficult to estimate pricing, particularly in non-Microsoft shops.

12

PROS: Snowflake doesn’t try to be something it isn’t. It doesn’t generate visualizations for the user beyond its basic administration visualizations.

CONS: For better or for worse, Snowflake needs to be integrated with a visualization tool like Power BI or Tableau.

Data Visualization & Dashboard Creation: 0/10

Data Blending, Preparation, & Automation: 9/10

PROS: Snowflake is basically a cloud-based reporting data mart with a super easy to use interface. Both experienced SQL admins and novice SQL developers will pick up the tool easily. All expected data blending, preparation, and automation options are available.

CONS: Snowflake is a scripting tool and so knowledge of SQL is a prerequisite.

Data Storage / Reporting Data Marts: 10/10

PROS: Snowflake is the best data storage tool for reporting data mart on the market today. Though in our opinion it doesn’t do anything revolutionary compared to other cloud products, its ease of use, user interface, and support model set it apart as the clear front runner.

CONS: Snowflake won’t, and shouldn’t, replace your enterprise data warehouse or lake. Snowflake is more purpose-built for supporting the critical reporting needs, rather than storing the entirety of your company’s data.

Advanced Analytics: 5/10

PROS: Snowflake itself isn’t really isn’t an analytics tool. However, you can think of it as them horsepower behind the analytics. The scaling of the warehouses behind the scenes makes it easy to prioritize and optimize the business-critical analytics while deprioritizing things that are less relevant.

CONS: Analytics are not built into the tool itself. This may require another tool, which can be expensive, in order for you to accomplish your analytics goals.

Pricing: 7/10

PROS: The Snowflake pricing is variable. It depends solely on how much you use the platform and how much horsepower you give it.

CONS: Because of its variability, it is easy to overspend if you are not careful.

13

How to know what is right for youWe have provided a lot of information about the tools available, but how do you know what to choose?

Our recommendations are to start small. Prioritize the biggest gaps you have in your company.

If automation is more important than visualization, start with Alteryx. If customization is most

important, start with Tableau. If a hybrid between the two is needed, Power BI is likely a good option.

Consider also that a reporting data warehouse, such as Snowflake, is often a universal need for

supporting any analytics toolset.

Our recommendations are to start small. Prioritize the biggest gaps you have in your company.

Successful implementation and adoption

05

15

McKinsey&Co found that organizations perform better when they have employees who understand data concepts

very well or completely. Understanding data concepts can be simply defined as data literacy. Moving toward a culture

of data literacy means organizations shift efforts from having only a small team of data specialists toward enabling

the capabilities of the entire workforce. By investing in tools and training, organizations make teams self-sufficient

and increase the opportunity for breakthroughs using the ever-growing sources of data.

Companies who are struggling with driving a culture of analytics often rely on online-learning platforms,

workshops, or executive-level programs. However, these quick-fix training solutions won’t transform an

organization. Instead of hasty training programs, adopt a holistic, cohesive training program that can scale with the

needs of the organization. Defined as an analytics center-of-excellence,these trainings drive cultural change and

provide all employees with the ability to collaborate and deliver repeatable solutions.

Analytics centers of excellence help scale adoption within organizations by retooling their existing workforce.

These centers establish common protocols and direct employees around data governance, data preparation, data

visualization, and other analytic processes. Effective centers enable their workforce to learn by doing.

To understand the extent of data literacy across an organization it must track performance on both the learning

and usage of tools. By knowing who is and is not adopting tools and learning, leaders can shift effort as they go

through the analytics transformation.

One limiting factor of analytics centers-of-excellence are high-level users who can both develop high-level solutions

and train and support learning of these tools.

When searching for the right data analytics partner, choose one that works with your organization to establish

and enhance communities through cohort-style training. By training in cohorts, peers are held accountable to the

progress of the group. And with industry-recognized trainers any organization will increase the capability of what

can be done with their analytics toolkit.

Data analytics training methods

06

17

Online instructor-led courses

With online, instructor-led courses, professionals have the flexibility to learn on their schedules

and at their pace. The challenge with online learning is the lack of support and accountability. The

instructor-led portion of the online course is typically just a person who recorded the videos for

the session, and not someone you can interact with on your schedule. If you get stuck, you often

have to look something up on your own–-and doing that will likely take you to a different site!

In-person workshops

With in-person workshops, you are supposed to get the opportunity to sit with peers and get

a boost of knowledge over a short period of time. The reality is that instruction is not always

consistent. Some trainers read directly out of a book, or worse: they sit there while you follow

instructions in a manual!

There are only a handful of training options for professionals: take an “instructor-led” online course, go to a 2-or 3-day workshop, or take a college course. Each of these options come with their own unique challenges.

18

For a workshop to be successful you need to have a strong trainer. In addition, as the participant, you often

need to have the perfect amount of knowledge and expertise going into the training. Often the examples in the

workshops are not tailored to your experiences, and that can make retention of information much more difficult.

When it comes to the psychology of learning, we know that adults learn through exposure to content over several

weeks, not days. We know that adults learn when examples are contextually relevant. We know that adults,

while they do not need hand-holding, often need occasional support where they can work with a coach on their

schedule. And we know that even as adults, we still need a little accountability in our lives.

Instead analytics professionals should consider other options to learn the top tools for their careers. Ideally

courses should be designed so that professionals aren’t overwhelmed at any given time and are extremely flexible

to anyone’s schedule-–even as professional and personal life shifts.

This learning needs a coach that can support when someone gets stuck, but isn’t constantly looking over their

shoulder. Ideally the coach is an industry-recognized professionals who understands the tools and technologies

they are learning. The best courses always end with capstone projects what allows an indidividual to wor with data

that is relevant to their jobs and careers.

When it comes to the psychology of learning, we know that adults learn through exposure to content over several weeks, not days.

Data analytics certifications

07

20

As organizations continue to invest in their teams it’s become increasingly important to track their employees growth in tools, technology, and processes. By tracking this information leaders know which teams have the skills to help transform organizations using data, and which are lagging behind.

As organizations continue to invest in their teams it’s become increasingly important to track their

employees growth in tools, technology, and processes. By tracking this information leaders know

which teams have the skills to help transform organizations using data, and which are lagging

behind.

When organizations use badges correctly, they provide additional benefits. As learners earn

badges they increase their confidence with their new-found skills. This increases morale and

results in higher employee retention. In a study by the O.C. Tanner Institute, they found retention

to be 26% higher due to use of learning badges. That same study showed that employees earning

the badges also tend to be 34% more productive. This means by using digital learning badges

organizations can increase their visibility to top performers.

As an additional benefit to employers, as employees share these badges it increases the digital

brand of the organization. Consider IBM’s adoption of digital learning badges, which included

adding badges directly into their staff directory: after implementing they saw a 57% increase in

certifications, and a 694% increase in internal training completion rates.

As learners earn badges they increase their confidence.

Retention is

26% higher due to use of

learning badges.

~ O.C. Tanner Institute

21

While the benefits are clear, for successful adoption to occur organizations should consider:

Increasing Communication For successful adoption, organizations must communicate consistently

and regularly about the opportunity. This communication should be

tailored to the audience. This means leaders should have different

communication than employees.

Increasing transparency

Whether is on a centralized site or on an individuals company profile

page, it should be extremely easy to see who has–and has not–earned a

digital learning badge. This makes it easy to identify employees for

intra-company hires. Increased transparency also motivates leaders to

drive their team through the learning process.

Make it meaningful

If you can tie your credentials to bonuses or promotions it makes it

extremely meaningful for employees. But when that option isn’t available,

it can be used for product owners to set and monitor access to data or

platforms so that data are used in an appropriate manner.

1

2

3

22

As you begin to implement

Data Coach is dedicated to providing you with the tools you need to achieve your

professional goals. It can be hard to articulate your skills in a series of bullet points

on a resume. That’s why we have partnered with Acclaim to provide you with

a web-enabled, verifiable version of your Data Coach badges and certifications

through a digital badging platform.

During the Data Coach program you will choose a project that you complete using

your own data. The goal is to build your project to meet measurable criteria that

demonstrate your understanding of the features and functionality of the tool.

At strategic points throughout the program you will work with a coach to get

feedback on your project. Once you finish your project and satisfy the criteria you

will be awarded your badge!

Once awarded, you’ll be able to manage your badges through Acclaim, then share your achievement wherever you’d like online:

Over email Embed on your personal websiteLinkedIn Facebook Twitter

Organizations are acheiving success is through custom badges or credentials.

While there are many platforms

the most popular is Acclaim.

Making smarter, faster, data-driven decisions

08

24

We scale and accelerate analytics by training people on how to be better analysts.

We’ve supported talent development for thousands of individuals, across Fortune 500

companies and some of the most innovative startups. Data Coach for Business pairs

expert courses with tools designed to support learning and development within an

organization of any size.

We are a group of seasoned analytics experts who are passionate about teaching others

how to become savvy with technology. We are experienced in coaching users at all levels

of an organization on how to make data approachable, consumable and insightful.

We provide opportunities for the cohort members to chat with the group or individuals

through the Data Coach Slack application. This way you can support and learn side-by-

side with peers.

Your coach is available digitally most hours of the day to help facilitate your learning.

Your coach will meet with you once-a-week for a thirty minute one-on-one, hold weekly

office hours, and do live digital training each week.

With the quality of modern data and analytics tools on the market today, investing in technology is easy. What’s hard is seeing a return on that investment.

Companies often overlook what’s the most important factor: the human element.

The future of data visualization and data analytics

09

26

Emphasis on quality training programs

With lots of learning programs available online, organizations are focusing on delivering high-quality content that

will deliver the largest bang-for-the buck.

Internal certifications drive learning and promotions

As organizations grow quality, internal training programs they will establish internal certifications that help

solidify data governance processes and enhance expertise.

Analysts will get closer to sources of data

Organizations continue to push for data democratization while working with more data. As the number of data

sources increases for an organization, those sources will be more difficult to maintain governance around.

Rather than taking a tight grip of these data pipelines and loading and transforming this data, organizations will

instead free their employees to interact at the data source and perform the necessary analysis, thus reducing

the time to action. With the advent of tools like FiveTran, Snowflake, and Alteryx, any employee can access data

sources, move the data into a scalable cloud platform, and prepare data for visual analysis.

Our team at Data Coach works with organizations small and large ranging from school districts to leaders of centers of excellence at the world’s largest companies. Our team has consolidated insights from our clients to summarize the future trends in analytics and data visualization that we are seeing.

27

As organizations grow quality, internal training programs they will establish internal certifications that help solidify data governance processes and enhance expertise.

Gold standard data sources

By freeing employees to focus on everyday analysis that happens within an organization, analytics specialists

can focus on building focused “gold-certified” data sources that act as a source of truth for executives. Over

the past decade, data teams have focused on making every data source a source-of-truth. This is impossible

to maintain. By focusing on data sources that are used for executives, these data teams can retain a level of

truth in data without becoming overwhelmed in governance.

Dashboards as product

Historically, business intelligence reports have had a very low level of polish. With the push for gold-standard

data sources, gold-level dashboards will take on a standard that requires a level of quality that resembles

a top-notch web application. This will require employees to have skills not only in the analysis, but also in

effective dashboard design.

©2020 Analytics and Data Visualization All Rights Reserved.

DATACOACH.COM