Data Storytelling With Excel – Episode I

This is the first of series I plan to do on how you can use Excel for pretty data visuals and data storytelling.

In this video I build a very basic Excel dashboard using data storytelling principles including:

  • Focusing audience attention
  • Removing clutter
  • Include insights and takeaways right into the charts

In this dashboard I use a 100% stacked area chart to visualize all users to a website and users who made a purchase on the website. The line where these metrics intersect on the area chart is the trend and the rate Made a Purchase/All Users. This visualization aims to show what percentage of the users who visit the website make a purchase.
There are a number of ways to visualize this, but I chose a 100% stacked area chart so I can focus on the rate. In this video the blue color on the area chart is for visitors who made a purchase on the website and the orange color is for all users.
Using a text widget I will include a clarification of what the white line between these two chart components is. This insight will also focus the audience attention on the rate which is what I aim with this visual.

The second chart in this dashboard is a simple line chart with annotations. It visualizes the trend for revenue made from website purchases. To be easier to understand the real numbers behind the trend I am including data labels and removing the vertical axis to reduce clutter.
Using a text widget I will insert a takeaway insight onto the chart as well as further takeaways on the right hand of the chart.

This is a very basic Excel dashboard, there are a number of other ways to make data look pretty in Excel which I will be exploring in future series.

Digital Acquisition Report Infographic With Google Data Studio

*Disclaimer: For this video I have used data from the Google Merchandise Online Store Google Analytics which is publicly available!

Google Data Studio has many options to style and make your reports pretty. I have been experimenting building infographics with Google Data Studio and I came up with some really great looking visuals.

In this video I create a digital acquisition report for a website to show performance by digital marketing channel. I got inspired by an available infographic template in Venngage, which is a great data visualization tool.

It was super easy to recreate a similar style infographic with Google Data Studio and in this video I show some important features like:

– how to select multiple widgets
– how to group widgets
– how to create new filters
– how to use background shapes and color

What I came up with is a very colorful report.
If you look at it from a data storytelling perspective it is a little difficult to digest and to focus your audience attention.
But as an unusual style report that looks like an infographic I think it is pretty cool.

Digital Acquisition Report Infographic With Google Data Studio

Hope you enjoy it and get inspired to create your own infographic in Google Data Studio.

The Pie Chart Dilemma

A pie chart is used extensively in presentations and reports. I use pie charts very often in my monthly performance dashboards. Or to be fair I used to.

The Naughty Pie Chart

In some of my recent videos I have used pie and doughnut charts without putting much thought.

If you watch my eCommerce Dashboard video you would agree it is hard to understand the pie chart I used for eCommerce conversion rate by country. The main reason being that the data contains many categories (countries) (slices) and a pie chart is totally unsuitable for the visualization of this type of data.

In fact it’s ugly!

So lately I have been very cautious about how I use pie charts. Or even doughnut charts!

Against the Pie Chart

Regardless of their popularity I have heard many data visualization experts to speak against pie charts and recommend using bar charts instead. The argument is that pie charts are hard to read.

When pie chart slices are close in values it is hard to determine which one is bigger. If the values are not close in size you can only determine which one is bigger but not by how much.

With bar or column charts however the audience will compare the end points. And because the values (size) of the bars or columns start at a common baseline it is easier to understand their relative values.

Also as discussed above pie charts are not suitable when you are comparing the values of more than a couple of categories.

In the visual below the pie chart is used to visualize traffic by marketing channel to a certain website. In this example we have seven categories included in the chart, each represented by a slice in a different color.

pie chart

Looking at this chart I have to agree it is hard to analyze each channel’s performance as compared to the other channels.

Below is the same data visualized with a bar chart. And it does look much better! It tells a good story that is very easy to understand.

pie chart vs bar chart

In Defense of Pie Charts

But pie charts do serve their purpose!

In the example above we are looking at seven different categories. But if we are visualizing only two parts of a whole a pie chart doesn’t look this bad. In fact I think I prefer it over a chart with only two bars.

Here is an example of a data visual for new vs returning visitors to a website.

pie chart two values

vs:

 

pie chart vs bar chart two values 1

Another big reason I have stuck with pie charts for a long time is that they are designed to compare parts of a whole. Or in other words they are the visual that will show percentage data labels.

All of the data visualization tools I have worked with only pie charts will calculate the percentage of each part to a whole. Including Excel.

Unless your numbers are already calculated and your tables include percentage values you will not be able to assign a percentage data label to a bar chart. (correct me if I am wrong here!).

Final Thoughts

Your lust for pretty data might lead you away from pie charts. And that’s understandable. But they can often be more effective if you are comparing two categories (two slices) within a total.

Custom Dimensions in Google Analytics via Google Tag Manager – Blog Categories and Dates

In this video I will show you how to create custom dimensions in Google Analytics via Google Tag Manager. I am going to use my blog posts meta data, the category and the date to create these custom dimensions. Blog categories and blog post dates are not default dimensions in Google Analytics. With a custom dimension we will be able to report on categories collective sessions as well as compare sessions between newer and older posts.

I am going to use the CSS selectors of my blog posts meta data to create a DOM Element user defined variable in Google Tag Manager. I will then use these variables to create the custom dimensions.

One very important thing to remember is to use your already existing Google Analytics pageview tag in Google Tag Manager to insert the custom dimensions. If you create a new tag this will mess up with your data, specifically Bounce Rate and Pages per Session. (sound common sense, but I have seen this mistake happen)

I am also going to create a custom report in Google Analytics with the new custom dimensions, Have in mind that it will take a day or so to start seeing data.

Follow the video for a detailed demonstration of this process.

Book Review: Storytelling With Data

Book Review - Storytelling With Data

Usually when talking about data one would refer to number crunching and staring at spreadsheets. In recent years however with the growing popularity of infographics and web analytics data goes far beyond just numbers. It is a visual process that require creativity, correct use of colors, correct use of graphs and most importantly storytelling.

I work a lot with data related to digital marketing and web analytics. I analyze a lot of websites and their marketing performance, user engagement performance and conversion performance and visualizing data is a big part of what I do.

A few months ago I stumbled upon Storytelling with Data blog, a blog which has actually been around for quite some time so my bad I only discovered it recently. It was clear from the very beginning that I had a lot to learn about data visualization from the awesome author Cole Nussbaumer Knaflic.

And it didn’t take me long to get convinced to buy her book Storytelling with Data.

The book is divided into ten chapter each one teaching a different concept in data design with real life examples and case studies. What I found particularly useful was Cole’s lessons in:

  • how to choose the most appropriate type of graph for your data (always a challenge when I’m working on my reports)
  • how to carefully choose colors
  • how to direct your audiences attention to where you want it

But what is most impressive about this book is its dedication to teach us how to leverage the power of storytelling to help get the message across and make stick with our audience. Chapter Seven called Lessons in Storytelling deals specifically with that looking at examples of storytelling in plays, screenwriting and creative writing and how the concepts can be applied to data visualization.

The book uses some really basic examples but as I was reading through all the tips and suggestions in the book I was amazed to realize how many silly mistakes I was making in visualizing my data in the reports.

Storytelling with Data specifically uses presentations and PowerPoint decks as an example. I wish the book had touched upon how to apply all these concepts and lessons when you deal with dashboards and large reports. In my line of work I sometimes need to put several visuals in one page. This is my biggest challenge every month when I prepare report for clients – how can I draw their attention to where it is most needed and how can I tell a story when there are so many graphs and tables in one place.

Still this book is a great starting point and an excellent resource to help you with any data visualization challaneges you are facing. You will get a basic introduction to data graphics, will learn the importance of color and how to apply storytelling concepts into data visuals.

How to Build Tables in Google Data Studio

Google Data Studio offers the abillity to build pretty tables with a lot of options to highlight your most insightful data.

In this video I will build a table with two dimensions and two metrics. I will show how to use bars or a heatmap in one of your metric columns to highlight the highest value. In this example we look at a table with landing pages per channel, sessions and eCommerce conversion rate for each landing page. The table will be sorted by highest number of sessions and I will implement a heatmap on the eCommerce conversion rate column to highlight the page with the highest conversion rate.

Also in this video you will see the new data panel which was very recently released by Google Data Studio. This new panel makes easier and faster to change or add new dimensions and metrics to your widgets.

If you are just starting with Google Data Studio check out my detailed Google Data Studio guide

Also check out my Google Data Studio examples of an Organic traffic dashboard and an eCommerce dashboard

Google Data Studio Examples

I recently published two videos with Google Data Studio examples of an organic traffic dashboard and an eCommerce dashboard. These videos show all the main features of the Google Data Studio to build any type of report.

Google Data Studio Examples

These Google Data Studio examples use Google Analytics as a data source and demonstrate how to add scorecards and backgrounds, apply styling, apply a heatmap to a column in your tables, and many other features of Google Data Studio to help you start building your own dashboard

Google Data Studio Examples – Organic Traffic Dashboard

This dashboard is an example of a monthly report on Organic search performance with a fixed date range of last month.

Google Data Studio Examples – eCommerce Dashboard

This dashboard is an example of an eCommerce performance dashboard with a data control option to let your client or manager select their own desired date range.

If you are just starting with Google Data Studio, I have written a detailed Google Data Studio Guide to get you started.

Line Chart Examples For Data Storytelling

These line chart examples show how to use annotations to not only show a trend whether in traffic, conversions or user behavior on your website, but to tell the full story about performance with your data.

I use line charts a lot in my reports since they are one of the best way to report and visualize trends, whether it is a trend in traffic, conversions or user behavior on your website.

When the trend is consistently showing an increase or decrease or staying stale it might be a little easier to tell a story with your data with a simple line chart and couple of comments. But when a trend is fluctuating as a result of marketing activity a chart line might look confusing and the data story difficult to understand.

Below is a line chart showing the trend for social media traffic for 2017 for a certain website:

Line Chart Data Storytelling

Traffic from social media is heavily dependent on social media campaigns activity and trend has been fluctuating throughout the year.
In the line chart example I have simply visualized the numbers for traffic from Social media each month in 2017.

Upon exploring I understand that these fluctuation are resulting from advertising activity as well as success of the specific campaign.

In order to fully tell a story with the data and explain traffic increases and sharp drops an analyst might want to include annotation to explain each significant fluctuation in the trend.

Line Chart Examples With Annotations

Here is a line chart example of how I improved the visualization and insert explanation of why we are seeing big increases and decreases.

Annotated Line Chart Examples Data Storytelling new

It might make more sense than the first example however after looking at it I thought that the background color was not a good idea. I also thought there are options to improve the annotations and make the main takeaways more visible at first glance.

So I made a few changes:

Annotated Line Chart Examples Data Storytelling FINAL

I changed the background to white and did some styling to the annotations text. I also decided to remove the data points and the data labels. All this is to try and draw the attention to the traffic fluctuations and the main reasons for these fluctuations which is what I am trying tell with this data.

Create a Combo Chart in Google Data Studio

*Disclaimer: For this video I have used data from the Google Merchandise Store Google Analytics which is publicly available.

A combination chart visualizes data for several metrics to help you compare. In this video I show you how to create a combo chart in Google Data Studio. I am creating a trend chart for sessions, unique purchases, product adds to cart and revenue for a period of 13 months. In this video I also show how to style a chart and choose different colors for each metric.

If you are just starting with Google Data Studio, I have written a detailed Google Data Studio Guide to get you started.

How to Create Prettier Reports in Excel

While it is the data in any Excel reporting that matters, good styling, colorful charts and graphs will turn a report from raw data to a compelling story and be easier to digest and understand.

Here are a few basic tips to help you create prettier reports in Excel and tell a story with your data

How to make pretty excel reports - Finished report

Styling Excel Tables

1. Start your data table in B2

Tables in Excel are not very easy to digest. Data is best understood when visualized with graphs and chart. However there are data numbers that need to be presented in a table.
Moving the table to start from B2 gives a nice spacing between the top and left margins of your report. This spacing simply makes it easier to look at a table of data and easier to read and understand the data

2. Branding and Report Title
It might also make sense to start even further down in column B and insert logos of your agency, client or brand at the top of each sheet on your report. Underneath this branding fill one of the rows with a suitable branding color by selecting the entire row and then fill it with an appropriate color. You can begin your table of data one row below this filled row.
You can also insert report title and date.

3. Always use table formatting
Tables with data are easier to read if formatted properly. Excel has a great table formatting feature with some really good built in table formatting templates. You can also build your own table formatting style to use or you can manually format your tables. When formatting tables try to use a header row, lines between each row and rotate colors between rows.

4. Remove background noise by deleting the gridlines from the sheet. You can do this by going to the View tab and untick the Gridlines option. This makes the sheet background white without any distraction making it easy to focus on the data.

Styling Excel Charts

5. Align axis titles on the chart horizontally
One of the things I hate the most about Excel charts is when I have to tilt my head to read the axis titles.
There are many ways to avoid this. You can change the chart type or stretch the chart to fit in the titles.

6. Delete gridlines and legends with one data series
In order to tell a story with your data the reports need to be clean of any background noise or anything not useful on the charts. I personally love charts and graphs that have no gridlines. Gridlines might be useful with a line chart but any other chart is better without them. It is very easy to delete gridlines, simply click on one of the lines to select and hit delete.
Also if your chart has only one data series it is no use to keep a legend of it. Delete it to focus on the important story the chart has to tell.

Pretty Excel Reports - Don't tilt

vs this:

Pretty Excel Reports - don't tilt 1

7. Sort your data before charting
Make it easier to understand what data is saying by sorting largest to smallest numbers before charting your table.
It is way easier to visualize how numbers in your data relate if they are sorted beforehand.

Pretty Excel Reports - Sort before charting

vs this:

Pretty Excel Reports - Sort before charting 1

8. Color your tabs
If you have a report with data in several tabs you have the option to give colors to your tabs. You can choose a color by right clicking on the tab and choose “tab color” option and select the color of you choice.

Pretty Excel Reports - Color tabs