DM course 3 : Marketing Analytics in Practice

Module1
Lesson 1: We talked about achieving the GOAL/DREAM – which is impactful insight of data as an
analyst.

To achieve this we talked about a framework by David McCandless.

So basically has 4 parts

PlanCollectAnalyseReport Plan……

Lesson 2: Planning- to define objective 1st step and most important

Now there are categories of Objectives related to different domains, for eg

1|Page
 DM course 3 : Marketing Analytics in Practice

So the broader goal is Business Objective (goal) in domain such as revenue, volume, Profit and to
achieve that we define marketing Objectives and further media objective and campaign objective
within.

So as analyst we focus on marketing goals which brings 5 primary goals that is

1. Build awareness
2. Influence consideration
3. Improve the sales process
4. Reposition the brand
5. Increase loyalty
Stimulate demand is also a goal but that is obvious and can sometime be overwhelming and may
harm than helping so it should be used carefully.

The same objective could be understood if we look at CDJ, we ask questions at different stages of
the journey

The objective discussed above can easily be identified in the CDJ

2|Page
 DM course 3 : Marketing Analytics in Practice

Now once we have got the objective the final step is draw the layout for further, this could be done
by asking right set of questions like – what, How and why and then moving in them to find data from
different sources.

To understand this lets take an example

3|Page
 DM course 3 : Marketing Analytics in Practice

Module 2
After completing Module 2, you will be able to:
 Learn a framework that answers the question 'what is good dataviz?'
 Use of common elements of a chart to clearly communicate the insight we intend to convey
 Create inviting visualizations that highlight key information for a reader and ease
comprehension
 Pay careful attention to detail to ensure dataviz is effective and efficient in communicating
insight
 Learn rules to create dashboards that rock

Guiding Questions
As you complete the lectures and assignments, you will find answers to important guiding questions.
These questions should be the focus of your learning.
1. What are the elements of a good data visualization?
2. What is the proper way to use titles, subtitles, and labels in a chart?
3. What are the categories of contrast that can be used in a data visualization?
4. How should we choose fonts, colour, and other design elements to improve a data
visualization?
5. What are the guidelines for designing great dashboards? 

Lesson 1: Principles of Dataviz

Here we are trying to find out what is the best way for data visualization (dataviz), but the answer is
not that simple, dataviz is an important part of analytics and one mistake can change the entire
meaning.

In order to have correct dataviz we should have completed the previous process correctly i.e
PlanCollectAnalyse Report(dataviz) and also it is important to keep in mind the concept of
David McCandless which includes Story, Goal, information, visual form.

So for a good visual form it should have 3 elements according to Wong.

1) Clear meaning
2) Sophisticated use of contrast
3) Refined execution

4|Page
 DM course 3 : Marketing Analytics in Practice

John Kenny says- the best data visualization is one which provoke conversation, it has a point of
view, and people may disagree with it and may provoke implications, but it is never boring. Bad
data analysis is boring. Data when delivered with a point of view can be transformative.

Point

1) When creating a good visualization narrative, analyst should to rely upon frameworks such
as:
a) The 4-Step Analytic Planning Process
b) McCandless’s framework Successful Visualization framework
c) Wong's 3 elements of Good Visual Form.

Lesson 2: Sophisticated use of contrast

Here we understand how to use contrast for dataviz. The 4 types of contrasts are
1) Size contrast – diff sizes help us understand emphasis/ importance of word/
shapes/number/ phrase, etc.
2) Shape contrast – use of icons is best example, other can be simple shapes such as circle,
triangle, etc.
3) Colour contrast – diff shades of colour, diff colour, keep in mind for colour blinded people
4) Contrived contrast – anything other than above, such as arrows, notations (A’), etc.

Keep in mind while using contrast with big number with scale for measurement for ex-5 B YT views
which is 1/9 of total internet traffic (scale).

Another method to use contrast is to use of sketches, first draw rough sketch to get the rough idea
and then you can transfer that to a visual form.

Points
1) Analysts tend to purposefully use contrast for these reasons; a) draw the audience’s
attention, b) pull out meaning from data, and c) effectively communicate meaning to the
audience.

5|Page
 DM course 3 : Marketing Analytics in Practice

2) These Pre- attentive attributes(contrived "made up") are visual properties such as; callouts
and boxes that we notice without using conscious effort and help pull out data in order to
communicate meaning more effectively.

Lesson 3: Ensuring clear meaning

Here we look at how our dataviz can effectively convey our message to our audience by correctly
using headlines, subtitle, visual cues, direct labelling and annotations.

In subtitles, you need to avoid what is called as intellectually blank statements, which is a statement
that really doesn't give any kind of insight. It's something that says there are three reasons why we
would do this, right? That doesn't give any insight into the three reasons it just tells us that there are
three. So you need to be really clear on what that insight is and then use the subtitle as your
opportunity to communicate that. 

6|Page
 DM course 3 : Marketing Analytics in Practice

Another method is eliciting and evoking emotions. This is a high-risk, high-reward approach. If you
understand the way an audience feels about a topic and if you feel that you have a good
comprehension of that, you can elicit that emotion in your story.Challenges that if you miss or if
part of your audience doesn't feel the same way, you run the risk of isolating party their audience

Points
1) Using a legendary key tends to force the audience to move back and forth between relevant
information and the visual representation. Whereas; a direct label has the audience instantly
engaging with the information.
2) Visual cues are an additional element of good visual form because they reinforce clear
communication and draw audience’s attention to what’s important.

7|Page
 DM course 3 : Marketing Analytics in Practice

3) Annotations are akin to tools that direct attention to important areas of a chart.

Lesson 4: Vishual polish

What Professor Hartman labelled as Visual Polish is “careful attention to details that will ensure
dataviz is effective and efficient in communicating insight.”
Example- 2017 Oscars

Lesson 4: Dashboard
 The most successful dashboards are succinct, one page visual. More than 1 page
becomes a report.
 Dashboards are an effective communication channel if these common mistakes are
avoided;

1) Too Much Data, too little insights

2) Too Long, Didn't Read
3) Pretty but Empty Headed(no insights).

 Avinash Kaushik named 5 rules for building a successful dashboard;

1) Benchmark & segments - Always use benchmark/ scale to compare from and further
divide them in segments (using different colours).

2) Isolate a few critical metrics - Never put all the metric, use best 4-5, don’t go above
10 metric.

3) Include insights not just metrics – Include insights in written form to summerise and
recommend.

4) Power of a single page – Dashboard should be of 1 page, don’t take font size below
10 for metric and 12 for benchmark, goals, recommendations.

5) Churn which is stay relevant and be flexible – Don’t consider dashboard as

permanent, as business changes so does goals and dashboard. Remove metrics as
soon as they are not required. Keep small subset (~25%) for long term as 1 year and
churn the rest.

8|Page
 DM course 3 : Marketing Analytics in Practice

 With an increasing amount of data available to analysts, dashboards became complex and
ineffective when communicating insights to clients. Subsequently, Avinash Kaushik re-
imagined it to be 5 quadrants known as an Action Dashboard.

Here is another example of action dashboard

9|Page
 DM course 3 : Marketing Analytics in Practice

Module 3
After completing Module 3, you will be able to:
 Learn how to put together an effective plan for an analysis project
 Learn how to utilize various techniques to collect data
 Learn how to use various techniques to analyze data and find patterns
 Learn how to build a report that effectively communicates insights

Guiding Questions
As you complete the lectures and assignments, you will find answers to important guiding questions.
These questions should be the focus of your learning.

2) What is Bellabeat's most pressing marketing objective?

3) What key questions need to be answered to help Bellabeat achieve its objective?
4) What analysis techniques will help pull insights from the data we've collected?
5) How can we present our findings in a way that makes it easy for an audience to understand?

Lesson 1 : Plan
As Bellabeat is a small company with a specific niche, its biggest overarching problems is low brand
awareness. So once we set-up our plan for brand awareness and would be to collect this data from
various sources and change our plan if required.

10 | P a g e
 DM course 3 : Marketing Analytics in Practice

Lesson 2: Collect
From asking question(like above) we may get some insight to work on specific topic/area or more, in
more depth eg- in why- tech adaption rate over time, so we collect more dataset and carefully select
our tools to handle data according to requirement.

Lesson 3: Analyze
Ok so we have data now, it’s time to visualize the data through various forms to better understand
it. The visualization might not be perfect but it should give valuable insights for your understanding.

11 | P a g e
 DM course 3 : Marketing Analytics in Practice

 Analysis is meant to convey a story with information; it does not in and of itself present an
argument, though it can be used WITHIN one.
 Remember, this is a KEY concept: correlation does NOT imply causality.

Lesson 4: Report
As we learned in previous module that it is most important part. Use of right methods and to
engage audience not throw them off.
Lets look at first improvement using colours, visual cues and headings.

But still not so much good, now let’s use all the methods

12 | P a g e
 DM course 3 : Marketing Analytics in Practice

Here it is so here we have removed all lot of un-useful data and each and evey aspect of the visual
has its own reason let us see how.

Points

 The data is the reason a visual exists in the first place. Effort is made to make it as clear and
comprehensible to the reader as possible.

 Ambiguities should be avoided; information must be clear to the reader.

13 | P a g e