Exploring the ‘Power of Visuals’ Using SAP Analytics Cloud

itelligence helps your move to the cloud and creation of a culture of business innovation.

Author –  Tanisha Gupta, Senior Consultant, itelligence UK.

According to IBM, 2.5 quintillion bytes of data are created every single day, now take a minute to try and get your head around that number; it is HUGE!

As the world becomes more and more connected, the amount of data circulating our planet is increasing daily, and at an extraordinary rate. Whilst a fantastic organ, the human brain can only consume so much data at a time so how are humans going to be able extract the valuable insight that exists in all this data that is around us. The answer is “The Power of Visuals”.

Organisations strive to provide a better service, expand their footprint within their existing customer base and also diversify into new and emerging markets. Many are turning to “data” (both internal and external) to help them identify opportunities to do that. The key is to be able to effectively present the data to people in a way they can gain insights easily and quickly.

Data Analytics is the science of analysing raw data to draw conclusions, while visualisation, on the other hand, is the art of effectively communicating that data in simpler terms. With the massive growth in data, trying to gain insight and patterns from data when presented with a series of numbers can be difficult, however by combining analytics and visualisation together you present the consumers of the data with a solution that will help them to quickly identify trends and exceptions and focus in on areas that need attention.

There are multiple tools in the market that can help end-users with visualising their data, and one of them is SAP’s latest analytics offering – SAP Analytics Cloud (SAC). SAC provides end-users with multiple options to convert complex datasets into simple yet engaging charts, graphs etc. and in this blog I discuss visualisation best practices and how I have used them within SAC to help users gain insights quicker.

SAP Analytics Cloud

Over the past year, I have worked with several organisations to help them gain better and quicker insights from the vast and disparate data they have available by using visuals. In all cases, I have developed the solutions using SAC and what differentiates SAC from similar tools in the market is:

  • It is a native cloud product with integrated machine learning and augmented analytics capabilities.
  • It provides the option to connect to on-premise and cloud products both SAP & others.
  • It uses the power of the SAP HANA database internally.
  • It combines reporting, planning, application designing and also predictive & augmented analytics all under one umbrella thereby opening up enormous possibilities to make sense of your data.

You can read more about SAC here

Key Principles of Visualisation

Before delving straight into the development of the dashboards and visuals, we need to take time to understand what the user is hoping to achieve from the project and below are 3 tasks I always undertake prior to starting the development.

  • Define a clear purpose – It is essential to define a clear purpose. The elements that form your design should help answer real problems, be it tracking performance, monitoring customer behaviour or customer satisfaction. Don’t develop visual elements that do not have a clear purpose or objective.
  • Understand the audience – It is essential to know your target audience. The charts and graphs should be in line with their expertise and requirements so that they can use the output effectively.
  • Effective Requirements Gathering – As with all software projects, gathering the business requirements is a key task and there are several ways we can tackle this. At itelligence, we find that using design thinking techniques to capture reporting and visualisation requirements to be very useful as it puts the user at the centre of the design and the requirement. By following an iterative process of capturing, ideating and prototyping we find that the user buy-in to the solution increases as they feel it is their solution and not something that has been forced upon them and the final solution that is delivered may not be what the user initially thought they wanted but by having iterative playback sessions what they end up with is what they need.

Now that we understand our audience and their requirements we can then move onto the actual delivery of the project and the design of the visuals. Below are some of the fundamental principles I always consider when designing and developing data visualisations:

Colours – The colour palette that is used is very important as it immediately helps users to relate to the information. When used correctly, colours can help the users quickly identify critical elements within the data and can be used to group similar attributes within the data together. Using company colours is always a good starting point in the design process as users find it easy to relate to colours they are familiar with.

SAC provides multiple colour palettes for end-users to work with and allows users to create custom palettes as can be seen below.

 

Diagram 1. Custom colour palettes in SAC

Diagram 1. Custom colour palettes in SAC

  • Importance of sketching – Colour is a tool, but it should never be the starting point. Conceive and plan your and visualisations in black and white whenever possible. Then, apply colour conservatively with the primary objective of highlighting and enhancing comprehension. Here are two basic tips to keep in mind when assessing how to use colour in graphic representations of data.

Importance of grey – Grey is one of the most important colours in your palette because it offsets the colours around it.

Diagram 2. Image taken from google images

Diagram 2. Image taken from google images

Importance of white space in reading colours – If you were painting, you could merge colours in one seamless quilt as shown in the left diagram below. However, in designing, every colour you use must have a reason for being there and should be distinct. In many cases, colours need to be distinct since they will represent a unique value or attribute, and to help the viewer’s eye read them as distinct and clear, the colours need room to breathe. That is what white space can do for you.

Diagram 3. Image taken from google images

Diagram 3. Image taken from google images

Within SAC, colours play a key role in presenting and emphasising the data as can be seen from the map image below where we use colours to focus the consumer’s attention to the countries that require attention.

 

Diagram 4. Using colours to emphasise countries in SAC

Diagram 4.  Using colours to emphasise countries in SAC

  • Fonts – As a general rule of thumb, make important text within your report bigger. Consistency is key, for example, if you decide that all your main titles will be 30 font, you must ensure that all your main titles are precisely the same size. The following figure shows how the font sizes of various labels should vary in general for a design.

Diagram 5. Image taken from google images

Diagram 5. Image taken from google images

AVOID USING ALL CAPS. The era of social media and text messages have created a new set of rules for using text. Today, all caps indicates that someone is shouting, which usually evokes hostility and defensiveness. They are also harder to read and take up more space and chances are higher for readers to actually gloss over and ignore them. It is fine to use all caps for the dashboard title e.g. SALES DASHBOARD but do not use them for the rest of the dashboard component titles.

Looking at the SAC image above (Diagram 4), you can quickly identify the main title and the subtitles for each element of the visual by utilising different fonts.

  • Focus on exceptions and trends – Emphasising important information helps the user in spotting exceptions and trends in a visualisation. Colours and fonts play a significant role in this as our brain naturally group things by colour. A good design exploits the seeing and decreases the amount of thinking a user has to do.

Let’s try a little exercise using both data sets below –

Try to determine as quickly as possible, how many times the number 5 appears in the list below:

And now

Diagram 6. Image taken from google images

It is much easier to “see the wood from the trees” when things are emphasised as can be seen above.

SAC provides multiple options to highlight exceptions e.g. thresholds and conditional formatting. Using thresholds, one can define ranges for a number and assign corresponding colours to highlight the changes.

 

Diagram 7. Example of thresholds in SAC

Diagram 7. Example of thresholds in SAC

SAC can also create hierarchies and hierarchies are an ideal way of installing a “manage by exception” culture within an organisation. They allow end-users to identify any outliers in the data at a high level with the ability to drill down through the hierarchy to get to the low-level detail only for the area that was identified as an outlier.

 

Diagram 8. Example of date hierarchy in SAC

  • Coherence – It is vital to keep the design coherent, especially when dealing with large amounts of data. Grouping similar KPIs together helps the consumer assimilate information quickly.

In SAC, you can assign a colour to a measure to make it consistent throughout the story. This can be done once for each measure, and this reduces the time required to develop a story as you do not have to colour code the measure every time in a new chart.

Diagram 9. Assigning a colour to a measure in SAC

Diagram 9. Assigning a colour to a measure in SAC

  • Use charts wisely – It is crucial to choose your chart type carefully. Using the correct chart type to convey the data is critical and a great rule to follow is to make sure that the chart type selected delivers insight, and it is not chosen because it looks good. A chart should always help users gain insight from data quickly. Charts should always be easy to comprehend; otherwise, they are not fit for purpose.

For example, charts, where the axis do not start at zero (0), can be confusing and can often lead to incorrect decisions being taken, as the users have not interpreted the data correctly.

One piece of advice I would give is to avoid using 3D charts as they distort the information and do not represent it accurately. Also, avoid using pie charts, as it is sometimes difficult to see differences between slices when the variances are small.

SAC provides end-users with multiple chart options and sometimes “less is more” and just providing a simple numeric point chart with variances or ranking data can provide powerful insights. Both variance analysis and ranking are simple techniques in SAC, which add a further new perspective to the presented information. Below are examples of both visuals.

 

Diagram 10. Examples of charts with variances and ranking in SAC

  • Location – Where you place information in your dashboard is directly proportional to how quickly a human brain will interpret it and attach required importance to it. On average people read information from left to right (Middle Eastern countries are exceptions). Below is a high-level layout to present information in your dashboard.

Diagram 11. Image taken from google images

Diagram 11. Image taken from google images

The infographic below is something I developed in SAC, and the location of the image and KPIs was a key part of how the consumer would read the data, with the main KPI’s starting in the top left and going from left to right.

Diagram 12. Importance of layout using infographics in SAC

Diagram 12. Importance of layout using infographics in SAC

  • Design Iteratively – Do not wait until your requirements are 100% understood. Visualisation requires an understanding of the visual concept your end-users have conceived and your ability to understand their optical needs. This is a tricky thing to accomplish, so do not wait to try to ensure you have a 100% agreement between you and your end-users. Get a big chunk of the requirements then start designing wireframes and prototypes, elicit feedback in an interactive setting, and revise your visual with direct interaction with your end-user.

Now that we have looked at the key visualisation principles, I would like to share with you some other cool features within SAC. As mentioned previously, SAC has in-built machine learning and predictive capabilities, which not only sets SAC apart from any other tool in the market but also helps with achieving improved visualisations in your story.

  • Predictive Forecasting: Predictive forecasting takes into account past trends, values, cycles and fluctuations in the data and uses this to make future predictions. There are several forecast options available as can be seen in the screenshot below. You can read more about it here https://www.sapanalytics.cloud/resources-predictive-forecasting/

Diagram 13. Predictive forecasting in SAC

Diagram 13. Predictive forecasting in SAC

  • Smart Insights: The smart features within SAP Analytics Cloud utilise the machine learning and predictive capabilities that SAP Analytics Cloud has to offer. Smart insights will analyse a specific element whether that be a KPI or a chart and tell you what the key drivers are behind the data you are looking at and you can see this from the commentary under the graph in the screenshot below.

 

Diagram 14. Smart Insights in SAC

Diagram 14. Smart Insights in SAC

  • Smart Discovery: If you are trying to understand which KPIs to develop or focus in on, Smart Discovery can build a list of key influencers along with multiple KPIs behind the scenes for you. From there, you can analyse and select the KPIs relevant for your business to create your visualisation, making sure they are in line with the best practices discussed below. You can read more about it here.

https://www.sapanalytics.cloud/resources-smart-discovery/

https://www.sapanalytics.cloud/resources-smart-insights-vs-smart-discovery/

 

Conclusion

The American statistician Edward Rolfe Tufte who is noted for his writings on information design and as a pioneer in the field of data visualisation pointed out, “the essential test of design is how well it assists the understanding of the content, not how stylish it is.” The goal is to enhance the data through design, not draw attention to the design itself.

The days of batch processing and making “finger in the air” decisions are no longer acceptable. Customers and partners expect answers at the speed they are coming up with the questions and data plays a key role in ensuring that informed, correct and real-time answers are being provided. With the huge increase in data, we need to rethink how we extract insights from the data to ensure we are providing informed and correct answers.

Providing decision makers with rows and rows of numbers in a spreadsheet is no longer sustainable as trying to spot trends and exceptions from this will and is becoming more and more difficult. This is where the use of visualisations and telling a story of your data is becoming important as it helps the consumers of the data focus in on areas of importance.

The next time you are tasked with developing a report, take a moment to think if the use of visualisations in the report / dashboard will make it easier and quicker for the consumers to gain the insight required to help make those real-time and informed decisions.

itelligence frequently run visualisation best practice workshops for our customers and if this is something you think would be of benefit to your organisation then please feel free to contact itelligence

 

 

 

Similar posts

Read more
Read more
Modern data platform
Read more
SAP Lumira High Quality Data Visualisation
Read more
Read more
Read more

Contact
Contact

For any questions please feel free to contact us.