Archives‎ > ‎Interesting Articles‎ > ‎

Effective presentation of business intelligence data

posted Jun 16, 2013, 5:04 AM by Leroy Dyer   [ updated Apr 21, 2016, 5:18 PM ]

 

Currently there are many companies and institutions presenting data and making enquires and investment into Business intelligence. Many companies are realising the potential of analysing big data and representing and presenting such data. Analysing web market trends, web usage data, medical data, customer sales, market rates, customer segmentation and trends, patterns can be identified and insights created. The presentation of business intelligence attempts to address this ideal. The way in which, results are displayed becomes a key factor so that the people reading the results can quickly and correctly identify the insights being projected. Business intelligence is a relatively new methodology although data analysis is not a new methodology to businesses. In the presentation of information generated by data, business intelligence has become a vastly growing field. This precise task is handled in many different ways. Charts and graphs are used to present data in a visual way, this enables insights to be easily visualized and shared with the appropriate audience. There are many software solutions available to businesses, we shall attempt to cover a selection of these software solutions and compare the types of presentation capabilities. At this present moment there are no specific methodologies in relation to the presentation of Business intelligence but some whitepapers have been written and approaches and methodologies are being discussed.

Data analytics, has vastly improved over the last ten years and previous methods of presenting statistical data has been unable to cope with the increasing demand and growth of databases and marketing information. Such databases have become data warehouses which capture vast amounts of analytical data. Mapping this data, analysing and effectively presenting multidimensional, data collected over time, categorized by location and other methods has become a growing concern for companies as data contains valuable information about customers and markets. This can have a great effect on sales and profits, therefore providing mechanisms to react to change effectively. Predicting market change, identifying valuable products or customers and even monitoring a company’s objectives and performance targets and achievements, has become a valuable asset. 

 

 

 


 

Which ways can we present data?

There are many different charts and graph types to present data, such as;

Pie charts: used for showing segments of the whole.

Bar charts: used to show data over time, (not good for small changes)

Line graphs: used to show multiple items in a group / comparing changes within the group. Useful for (small changes)

scatter plots: useful for comparing differences between two different things: These traditional charts can be generated by various online and offline creation tools a blog offered at Hongkiat.com give a list of 22 chart generator software’s available. (Listed below)

 

·         Rich chart live: (ric)

·         DIY chart : (diy)

·         Online chart generator : (cha1)

·         Chartel.net : (cha6)

·         ChartGo : (cha5)

·         Create a graph : (cre2)

·         JS charts : (cre1)

·         Pie chart tool : (nce)

·         Piecolor : Hohli charts: (htt)

·         Css chart generator : (css)

·         Chart part : (cha4)

·         Chart maker : (cha3)

·         Google chart tools : (cha2)

·         amCharts visual editor : (amc)

·         pie chart maker : (cre)

·         ChartGizmo : (cha)

·         Online chart tool : (Onl)

·         OWTChart Generator : (Hig)

·         HighCharts : (Hig)

·         ICharts : (ICh)

 

These third party tools offer traditional chart making facilities, yet for business intelligence and large data sets, more dimensions and measures are required. Greater software is required. Dealing with large datasets the information presented is often taken from multiple data sources; this data is not easily handled. This has given rise to specialists in business information technology company solutions.

 

·         SAP AG, SAS, ORACLE, IBM, MICROSOFT

 

These are the top five vendors in business intelligence software, although there are others making headway and providing new innovations in the presentation of data;

 

·         Tableau Business Intelligence: (tab) : dashboard creation:

·         QlickTech: (qli): A free software business chart making software.

·         WebFOCUS: (inf): Web enterprise integration software.

·         TIBCO Spotfire: (spo) : Information architect, data mining, predictive. Analytics

 

Currently there are many database management systems adding this functionality to current solutions.

 

Locational Intelligence

This offers annalists an additional level of analytic capability the ability to have multiple layers of information is a key requirement to the business intelligence process as organization and classification of data enables a greater understanding of unstructured data.

 

“Integration of Location (GIS) and standard BI platforms brings LI to greater usefulness by making it available as an option to anyone who is familiar with the more readily-available BI solutions, and without the need to master new concepts or a new user interface. Spatial relationships also greatly enhance many of the details commonly reported by BI systems, providing an added level of analysis that is useful in viewing and assessing trends (and existing data types).” (sau)

 

A recent research report – (sau) discusses locational intelligence, with advanced mapping capabilities presenting data can show geographical information in innovative and informing ways. Trends can be mapped and visualizations can reflect data down to street level, for customer trends and marketing this is a useful way to determine current customer movement and trends.

The ability to access information “on the move” has become a rapidly growing area of business intelligence, accessing dashboards, and scorecards keeping in contact with real-time information and information sharing has become of great importance to data driven companies. This area of business application development is a fast growing area of business information presentation. The simplification of dashboards is becoming increasingly important. The benefits of mobile business intelligence are significant and are confirmed In a TDWI best practices report – (MOB)

By deploying dashboards in the cloud making the dashboard or scorecard available to users immediately has also given application developers an efficient way to connect to and collaborate.

The creation and deployment of dashboards and scorecards is also a key factor in presenting business intelligence data. Operational / Strategic often referred to Executive and Analytical are the main types of dashboards created. Each dashboard type has a level of detail or summarization.

These dashboards are often used to monitor data and traffic and real-time information. Operational dashboards have added considerations such as refresh rate, as the refresh interval can affect performance of the dashboard.  

These dashboards often contain key performance indicators tracking costs, sales and performance over time. Annual, quarterly and monthly this level of summarization is generally all that is required for such dashboards.

 

These dashboards also offer additional drill down functionality, enabling deeper insight into the data being analysed

 

When designing dashboards it is important to select the right type of dashboard, as adding functionality where it is not needed clutters the dashboard. Anything selected to be placed in a dashboard must have relevance to the insight being displayed. Selecting the correct chart is also important. Many charts can be created for various types of data presenting and selecting the KPI or metrics required is often important for the insights being shown or projected. Tableau software now has 23 different chart types for dashboard design. This is bringing the tableau software to the current market leader in dashboard technologies. Oracle and Microsoft are also leaders in the development of business Intelligence dashboards. Microsoft Excel has been used to also create pivot charts which have also been widely used. A useful dashboard should be tailored to the employee or users’ needs and requirements. By using caption and headings to identify charts is also an important factor as concerns with placing too many charts on a dashboard often comes with the added sacrifice of “explanation”.

 

I believe yes; dashboards enable the necessary business stakeholders easily identify information; with traditional database queries and spread sheet analysis often information is not easily identified, and easily missed. Insight are difficult to determine and relationships not always apparent. Using software such as Tableau OLAP cubes can be visualized more effectively. Dashboards can contain maps, charts, tabulated data, and word clouds and other visualizations; the human mind reacts well to such stimulus. The deployment of the dashboards is relatively simple, the functionality and metrics on a dashboard give the company an insight into the data being presented. These interactive charts have enabled business and data modelling to become a fore runner in the big data industry and also in the management and analysing of sales and customer data collected from online transactional data collection. This makes them invaluable to data driven business models.

Dashboards are being commonly used in many forms from gaming to flight statuses. Poker sites use dashboards to present gamers with statistics about games and present winning statistics even offering comparisons with other player’s statistics. Whereas websites offering flight-stats offer present flight data and seat statuses. Such dashboards could be classed as consumer business intelligence. These types of “consumer” dashboards are used to present data to web users or gamers seamlessly over different forms of platforms from gaming consoles to mobile applications. The consumer often is unaware that they are using a dashboard or scorecard. Consumers are being trained in the use of business intelligence dashboards via such portals. It would seem that business intelligence and presenting data has become common place.

With the increasing development in analytical modelling software, companies are now employing advanced visual modelling algorithms and techniques. By visualizing data, companies can visualize their data in various unseen dimensions. Customers can be represented graphically, even geographically or even as clusters which can be segmented by using advanced regression techniques such as , separating customers by Recency, Frequency and Monetary. This gives us the ability to determine valuable customers. The valuable customer segment can then be identified. Their similarities examined, this can then be used to determine business strategies and valued customer insight. The ability to display 3d or interrogate views is of a paramount importance in the increase of these advance business techniques and algorithms. Visual modelling has now become a major driving force in the development of customer driven businesses and has impacted many companies sales and allowed them to achieve an increase in specific target marketing strategies.

Insight driven graphical representation of business data and operational data have already provided businesses with valuable customer insight, market information and predictive analysis. The ability to monitor market changes visually has provided growth opportunities to companies proving its value in the business marketplace. Amazon have used transactional and web mined data to identify customer buying strategies, such data is used in developing customer insights and targeting mechanisms. This data can also be made available in shared data ware housing knowledge bases ready for advanced analytics and visual modelling techniques.

Development Lifecycle

In developing a business intelligence project or solution, there are some different perspectives, whereas some believe that there are no specific methodologies one would say that a software development approach should be taken, such as a waterfall model or even an agile approach.

Application development model

PLAN

 

 

 

 

 

 

ANALYSE

 

 

 

 

 

 

DESIGN

 

 

 

 

 

 

BUILD

 

 

 

 

 

 

DEPLOY

 

 

 

 

 

 

REVIEW

 

 

 

 

 

 

 

The above illustrates the waterfall approach to a business intelligence project; this can also be applied to dashboard development yet perhaps not to the actual design of the dashboard.

This model can contain multiple iterations and stages revisited if necessary.  

PLANNING

In planning a dashboard requirements gathering should be a main part of the planning stage, identifying the users and uses of the dashboard to be created, identifying the insight required to be shown.  These insights play an important role in the overall selection of the correct key performance indicators and metrics. Displaying and applying these metric are key to the dashboard becoming a success.

AYNALYSE

The data used within a dashboard is the key driving influence behind the dashboard being designed. This stage of analysis of results is a key issue. The data being displayed may be able to be analysed by using a data modelling technique such as Recency, frequency and monetary valuable insight can be shown. There are various types of dashboards for various levels of operations ranging from performance dashboards to strategic dashboards and real-time monitoring dashboards are also currently deployed across multiple platforms, from gaming to mobile business analytics applications.

DESIGN

In the design of a business intelligence dashboard, an approach called the “Three threes” is often commonly used. THREE APPLICATIONS, THREE LAYERS, and THREE TYPES: By integrating the application of the dashboard for multiple application types i.e.: monitoring, analysis and management this gives the dashboard a use which is more than just displaying data, by adding this functionality to the dashboard, we are able to create a dashboard that can also look at a snapshot in the current area to which it is being applied. By adding multiple dimensions to a dashboard giving the user the ability to drill down and delve into the data to take a closer or more directed look at the data, or even ROLL UP? This gives us a grand overview and perspective on the data.  By determining the role of the dashboard to be applied, the correct type of dashboard can be created.

BUILD

In the building of a dashboard there are many products supporting dashboard creation.  This needs to be evaluated according to the data or deployment strategy or even tools available to the project. Connecting a dashboard to existing technologies can be a tricky process and not all of the vendors of business intelligence products are able to accomplish this goal. The selection process also can be limited by budget, so understanding the business requirements can also determine which software vendor is used.

Deploy

The deployment of the dashboard is also a major consideration, deploying in a cloud based environment, or web based environment, creating a standalone application? These are the questions which need to be answered. By determining the access requirements needed by the dashboard the correct decisions can be taken. With mobile BI a media rich or even an html5 environment can be created and deployed on a platform that can be accessed by modern smart phone technologies.

REVIEW

This stage is mainly testing and reviewing the requirement stages to ensure that all requirements have been met, exploring the functionality and metrics applied. By using spread sheet technologies to check the validity of results.

This methodology can be approached as a rigid or agile approach, iterations can be applied to each stage revisiting if required.  This approach is entirely as flexible as required. By taking a methodical approach to designing dashboards and business intelligent presentations a user and company’s goals can be achieved.

Whilst it could be said there are no specific methodologies for developing an effective presentation of business intelligence data. Various approaches and considerations and ideals remain essentially the same. Target audience, functionality, colour, insight, analytics’, metric and kpi’s all play key roles in the decisions taken when designing or presenting data. Taking an approach such as the software development lifecycle or agile approach or even a prince2 or scrum approach is highly advisable. In reflection when presenting BI the choices are limitless, personal choice target audience company ethos play a major part in building bi presentations. New charting types and tools are rapidly expanding the field of business intelligence. With the introduction of word clouds and tag clouds keyword metrics and search data and social website mining has opened up the field of displaying qualative as well as quantitive data. This will be exciting for market researchers and broader researchers as these types of analytics have been a manual task or even beyond the range of categorization and classification. It could be argued that there is no need of a modern methodology as taking a rigid approach to data can obscure insight generation and discovery; agile methodologies enable a wide scope for creativity and development.