How to Define Effective Test Scripts for Dashboards

test scripts
8/3/2016

Testing may be one of the most important steps in your project to deliver robust dashboards and reports. However, in my 20+ years as a Business Analytics consultant and coach, I have found that rarely do clients focus proper attention on testing. Most feel it’s a necessary evil. I will now share some tips on how to define effective test scripts for dashboards and reports.

The benefits you receive when you know how to define effective test scripts for dashboards include:

  •  Increased end user adoption
  • Decreased end user training time
  • Reduced project time
  • Reduced project costs
  • You look like a rock star!

Before we start, I want to disclose that Lodestar Solutions believes the best implementation methodology is an agile or Scrum approach. In an agile project, you will define user stories that help to define the testing process. Testing in agile is done incrementally as opposed to the end of the project. This results in the ability to fix errors faster and provide for improvements as you go. For more information on agile, check out our "Save Time With Agile Scrum Methodology For Business Analytics" blog.

For those of you not on an agile or a modified agile methodology, you may want to leverage sticky notes to jot down everything you need to test. I recommend starting this at the beginning. Have everyone jot down what will need to be tested on a sticky note. Next up...put it on a large flip chart or store them in one place. Additionally, you may want to color code them. For more information on leveraging sticky notes, check out my video and join our Analytics Coaching Club, AT NO COST, for more time saving techniques.

Test Scripts on Dashboards

We believe that there are four areas you need to write test scripts to ensure proper dashboard and report testing. We will assume the data coming into your data warehouse or other source systems has already been tested.

test scripts
  • First, you need test scripts to ensure the data moving from various sources like the warehouse into your BI tool is mapped correctly. We need to ensure that data is exactly as expected.
  • The second area requiring test scripts is security. This is to confirm users can only see what they are supposed to see.
  • Next, test scripts are needed for the report or dashboard functionality. Ensure they meet the needs of your end users. You will also test security again at this point in case something was set wrong in the premliminary security test scripts. Always ask what the end users will want to do next with the dashboard. This gives you an idea on what features you could add in a later release like drill through.
  • Finally, test scripts enable the system to test performance and scalability. Systems often function when hundreds of users are on at once.

Getting end users involved in the testing early is essential to delivering a solution they will love, hence increasing end user adoption. The trick is getting them to commit their time to explain that they will have influence in features so it meets and exceeds their needs.

Remember testing of Business Intelligence (DW/BI) applications is a little different than testing traditional transactional applications. It requires a data-centric testing approach. Check out “How to Conduct Effective Testing of Business Intelligence Applications” which provides some good information on challenges you may face. As part of the daily or weekly testing, we recommend you share the results with the involved stakeholders.

Why A Test Summary Report?

After you complete testing, generate a test summary report. A test summary report explains various details and activities about the testing performed for the project. In addition to generating this report, it can be given to the respective stakeholders like senior management, client, etc. If you would like more details on a test summary report, read “A Simple 12 Steps Guide to Write an Effective Test Summary Report” which is quite informative.

So what are you waiting for? You can start defining test scripts with sticky notes today! To view my brief video which shares a tip on “How to Define Effective Test Scripts for Dashboards” and to join our Analytics Coaching club for more timesaving tips, click HERE.

To join our clubs, simply click the “Register Now for More Time Savings” button under the video. You will then be able to choose which clubs you’d like to join. Then click the green “Purchase Now” button to the right. Please note that there is NO COST to join! Once you provide your information and click the sign up button, you will have immediate access to amazing videos to make you look like a rock star.

Finally, if you have direct questions about setting up testing scripts or would like to learn more about our recommended Business Analytics Agile Methodology (BAAM), contact us at Coaching@lodestarsolutions.com.

Choosing A Visualization For Your Dashboard

3/1/2016

A C-level executive recently handed a 42-page stack of tabular reports to me and said business dashboards would need to be a priority for him. This isn’t uncommon. In today’s analysis paralysis world, analytic individuals often mistake exploratory analysis for explanatory analysis. To move to a true bottom-line impacting analytic culture, you should move away from laborious reporting to quick glances at key metrics which impact strategy and tactical decisions. That is the main goal of business dashboards but choosing a visualization is just as important. You should be able to very quickly glance at the key metrics driving an operational area and know whether to stay the course or change actions for a more favorable outcome.

In a 30-minute business analytics coaching session with this CFO, we were able to boil down the 42 pages into 5 KPIs (Key Performance Indicators) and 5 trends he looks at regularly to adjust actions. However, one of his challenges was choosing a visualization. It’s our job to choose a visualization which allows the reader to quickly interpret data. We want to keep cognitive processing to a minimum. We don’t want them spending time thinking about, “How do I read this graph?” We want them to quickly determine what’s my next course of action.

I always recommend, “Storytelling with Data” by Cole Nussbaumer Knafilic. She masterfully breaks the process into 6 simple steps. Even if your not a creative person, this book will help you quickly and easily implement function, form, and design best practices. Here are a couple of tips to get you started: Simple Text If you only need to convey one or a couple of numbers, use simple text. In the Gross Profit example, we might be tempted to put the data into a bar chart, but that’s added information the reader would have to weed through when all they are really looking for is a single number telling them how off target they are.

Choosing a visualization
Business dashboards – Simple Text

Tables In our analysis paralysis world, we are tempted to put lists or tables on business dashboards. Tables are great for mixed audiences because they will just locate the row and column that is pertinent to them. However, just because the dashboarding tool “Workspace Advance” from Cognos has scrolling functionality for large tables, you should still think carefully before using it on a dashboard. Use tables for summarized data and allow the end user to drill-thorough to a more detailed report if they need to get into the weeds of the data. In the Table example, they summarized the information as a % of total year. This allows for a quick comparison on how they are doing versus other regions. The table is also drillable so they can drill down to the country for comparison in a particular region.

Choosing a visualization
Business Dashboards – Table

To take the ease of understanding one step further, in the heatmap table, I put conditional formatting on the data to draw the end users eyes to higher percentages.

Choosing a visualization
Business Dashboards – Heatmap table

Bar and Column Charts Bar or column charts are very common as they are common for a reason. We all know how to read them. Bar charts are great for categorical data, however, as we add more data points, each categorical data becomes harder to read. I recommend that if you have more than 3 data points for a category, consider using a point graph. You can see in the bar graph example below that trying to compare multiple regions against multiple years is challenging yet in the point graph, the data is concise and easy to interpret.

Choosing a visualization
Business Dashboards – Bar graph
Choosing a visualization
Business Dashboard – Point graph

Destination Dashboard If you are looking for a dashboarding workshop where you can learn about choosing a visualization, Lodestar’s 2-day “Destination Dashboard” workshop, which is offered periodically throughout the year, will cover visualization development in-depth. You will leave the workshop with a physical blueprint, a repeatable process you can teach your organization, and Cognos training on how to build in Cognos Workspace. To learn more about the next Destination Dashboard events, contact us at Coaching@LodestarSolutions.com.

X