Introduction and Purpose
The terms ‘Big Data’, ‘Data Science’ or ‘Data Analytics’ have becoming increasingly popular with numerous conferences and symposiums of late. Reports on trends in Data Analytics reinforce the directions of the Organisational Excellence Philosophy defined by the various worldwide Excellence Frameworks. In particular, the Category representing the Management System of Measurement, Information and Knowledge described in all frameworks.
My main points in this discussion are:
- Trends and useful ideas observed about Big Data and Data Analytics reinforce the ideas expressed in the Excellence Frameworks. The ideas from Data Analytics add definition to the Approach and Deployment to developing a system of Measurement, Information and Knowledge within an organisation
- Attempts to introduce ideas reflecting the Data Analytics trend are likely to fail without a clear enabling framework as found in the Excellence Models e.g., a lack of an architecture of measurement tied to strategy, a lack of commitment to studying variation emphasised in the Excellence Frameworks
Big Data – Data Analytic Trends
The MIT Sloan Management Review Research Report 2013 describes a growing belief that organisations are “…on the cusp of an analytics revolution…” that is having a transformational impact on organisations (Kiron etal 2013:2).
Organisations have more data than ever before at a volume unparalleled in history. The volume of data stored in digital format is now enormous made possible by the advancements in storage devices and the advent of the ‘cloud’. Data is big because of volume such as a trillion phone records and big because of a large number of categories or dimensions say a few trillion.
Court (2015) describes new tools and improved approaches that are emerging to aid the Data Analytics Value Stream in organisations.
- Analytics based software and service providers are providing models with specific uses that can be implemented quickly across many organisations with the common need.
- New self service tools are democratising the use of analysis into the hands of frontline users. These tools allow non statistical gurus to link data from multiple sources and apply predictive analysis. Visualisation tools are an example.
- It’s becoming easier to automate processes and decision making. Broader capture of real time data across processes is becoming easier. Automated analysis enables patterns and relationships to be seen quickly and decisions taken.
Example: Tree map view of a social network’s track selections from a streaming media service. Keahey (2013)
Characteristics of Data Innovators
Using survey methodology, Kiron etal (2013) classified organisations into three groups reflecting their maturity with Data Analytics and. Their observations are similar to other reports (Morgan McKinley White paper 2015, Court 2015, Manyika etal 2011). Kiron etal found a small group of organisations which they termed ‘Innovators’. A summary of their characteristics are as follows.
Valuing Data – Data is viewed as a core asset and is used to encourage new ways of thinking, to be inquisitive, to test things, learn quickly and challenge the status quo. Data is grouped in different ways to gain insights.
Talent and Capability is in evidence – Through recruitment and training there is widespread capability across the organisation. There is Analytics support for leaders and there is a parallel shift in power and resources towards those who demonstrate data- driven decisions.
Use more of their data more quickly – Tools and services are used to understand operations and customer behaviour at a granular level surfacing questions and improvements which underpins new waves of productivity. They collaborate more around their analytics. The cycle time to generate reports has been reduced dramatically with self service analysis.
Deliver Results from Effective Use of Data
- Patterns of consumer activity are being understood that was not possible before. So Services and Products are getting improved and improvised as they get used. A Food organisation examining data sets on segments based on demographics redefined the target market counter to prevailing views.
- Health Fund used member data to identify risks with pain medication usage causing risk of heart attack.
- Electricity organisation collected hourly data to help monitor and give feedback to users and to improve energy grid management
- Beverage organisation used an ordering algorithm that lowered rate of inventory out of stocks.
- Insurance organisation predicting severity of claims and modifying operational model
Williams (2015) in a paper presented at UTS summarised 5 ways Data Analytics has made a difference as follows:
- Transparency: exposing patterns and insights
- Experimentation: exposing variability and insights
- Segmentation: allows customised actions
- Decision Making: using insights and prediction
- Innovation: drive new opportunities, business models and services.
Issues Observed
Court (2015) reports that several years ago, the McKinsey Global Institute estimated that retailers could increase margins by 60% and the Healthcare sector reduce costs by 8%. However, what has transpired is that very few have been able to achieve significant impacts.
Kiron etal (2013) identified organisations that were ‘Challenged’ by Data Analytics or had not extracted the most value from their practices – ‘Practioners’. Their analysis of these organisations (as with other reports) have identified issues explaining why many organisations struggle with establishing an effective system of measurement and analysis. In summary the issues are:
- Getting senior managers to understand the potential for data analytics and to invest in tools and training as well as become practitioners themselves.
- Lack of quality of data and data governance – not reliable, adequate, accurate or timely
- Collecting a lot of data with no plan to use it – capturing data just in case!
- Not obtaining or building talent and capability at all levels
- Not tying disparate databases together – lack of collaboration across the silos where different functional groups have built their own data stores
- Front line managers and other users lack confidence that analytics will help decision making and so continue to trust their own intuition and beliefs.
- Core Processes are not measured adequately nor are opportunities for automation in measurement and analysis taken
- Even when real time analytics are available in an organisation, the cycle times of management processes and decision making don’t match the speed of the available information and so opportunities to deal with issues and improve are not taken
Lessons Learnt
Focus on Change Management – Don’t try to orchestrate change in all areas of decision making and processes – it’s too overwhelming. Focus on smaller scale areas where there is an important issue, data is available and an impact can be made – e.g., pricing, inventory, credit management, customer turnover. Consider which functions / processes would benefit most from improvement in measurement and analysis.
See and Improve the Measurement and Analysis Value Stream – Develop a systems view of gathering data, analysing the data and using the data just as you would with any other process in the organisation. Ensure that strategy drives the measurement architecture.
Job redesign. – Redefine roles to use analytics and develop capability. Build the capability requirement into all levels of workforce planning – roles and responsibilities, recruitment, training and development.
Build analytics into the culture. Recognise teams that generate insights and improvements through analytics. Provide ‘boot camps, for end users to learn tools, create a community of power users to support staff and leaders – show you value data. Encourage collaboration across the value stream in the understanding of performance and customers.
Organisational Excellence – A Clear Framework for Building a System of Measurement and Analysis that Drives Learning, Improvement and Knowledge
Effective practices and barriers reported as part of the recent literature on Data Analytics resonate with what the Organisational Excellence Frameworks and literature have been advocating long before the current emphasis.
At the heart of the Organisational Excellence philosophy is the drive towards continuous learning, innovation and improvement on the basis of gaining knowledge. The selection of what to measure, data storage and access, the analysis of data, decision making and knowledge management routines is a major element of the Organisational Excellence models. The conceptual basis for this area is Understanding Variation and Statistical Thinking – Numeracy Skills in organisations!! Research has shown that practices in this area and scores on organisational assessment on this area differentiates high performance organisations from others that struggle (Evans and Lindsay 2014).
Data Analytics Discussion Missing Emphasis on Measurement Architecture
The emerging characteristics of effective analytics practice are captured within the Excellence Frameworks dealing with Measurement, Information and Knowledge. The characteristics of Innovators provide further support for the content of the Excellence Frameworks and add more detail to the practices proposed in the Frameworks.
However, discussions on Big Data and Data Analytics rarely emphasise the need to establish the organisation’s measurement architecture which is emphasised in the Excellence Frameworks. Rather emphasis tends to be given to predictive analysis. Without establish a measurement system based on strategy that allows interlinking of organisational measures, the basis of asking predictive questions and pursuing predictive analysis lacks rationale. Therefore, a key step in maturity and innovation in Data Analytics is to establish the system of measures based on some strategic rationale for the organisation.
Kaplan and Norton (Financial Executive 2004:44)
Little Emphasis on Understanding Variation in Data Analytics Discussion
The focus on Measurement, Information and Knowledge stems from the founding thinkers like Shewhart, Deming and Juran. At the heart of the Excellence Philosophy is:
Implicit in this restatement of Deming’s view (Nunnally and McConnel 2007:44) is a need to understand Predictability or Stability and Capability. One of the significant omissions from much of the discussion about Data Analytics is the need to have the ability to understand Variation. This inherent capability is the cornerstone to understanding performance as a system and to determine whether improvements have been made.
Pattern of Stability and Variation over time has Changed
Pattern of Capability Variation over time has Changed
The characteristics associated with an effective approach to an organisational ‘System of Measurement and Analysis’ are summarised below. Jumping into Big Data and Data Analytic initiatives without the following characteristics built over time in an organisation is likely to result in wasted investment. I clearly understand that this takes time. However, this kind of investment has a natural return on investment for an organisation in terms of reduced costs and increased revenue and margin let alone market satisfaction and staff engagement.
A culture valuing learning, curiosity, understanding variation, openness, innovation and improved performance. This is what a culture that values data looks like! It is not simply having lots of data and analytical tools. Data will not be valued without a culture of learning.
A strategy driving a measurement infrastructure. Asking good questions given the strategy of the organisation and the challenges it faces drives the design of a measurement architecture that serves decision making throughout the organisation. Without purpose behind measurement ‘garbage in and garbage out’ is likely!
An analytics capability of different levels built overtime across the organisation. The capability rests on the ability to study patterns of variation in terms of stability and capability to predict performance and drive improvement that delivers measurable improvement that is sustained.
A deployment of practice that pervades all levels of the organisation. This means the majority (if not all) of employees. So the leadership team and all managers and supervisors become comfortable with a level of data analytics and interpretation. It also means that all teams have a level of capability as well. No part of the organisation is left out! All functions are involved in the use of data and analytics in some way.
Without the drive to interpret data, learning suffers, decision making on the basis of evidence suffers and so measurable improvement that is sustained suffers. Imbalance in decision making dominates since intuition, ‘gut feel’, ‘jump to solutions’ is what pervades the way decisions are made – particularly changes – from strategic change (new products-services, markets) to changes to the value stream.
An integration of data stores across the organisation. Instead of looking at performance measures as isolated elements there is a practice of systems thinking to interrelate measures (which starts with strategy). Analytics are used to help understand predictive relationships and patterns in data. This information is used to aid strategic decision making as well as decision making to aid value stream improvement. Instead of silos there is a holistic approach to the management of data including access that allows self serve analytics with appropriate software platforms.