Critical Success Factors of BI, Data & Analytics Implementations in the Fourth Industrial Revolution

About 5 years ago I wrote a paper on Critical Success Factors of Business Intelligence implementations. I have always been immensely passionate and fascinated by BI, Data & Analytics and keep a close watch on the direction these disciplines are heading. I was in the pursuit of finding the potion of just-the-right factors which, when applied correctly, would ensure the success (and continuity) of implementations that create and monetise information assets in organisations.

At the time, most evidence pointed at a multitude of unsuccessful projects in the space (with an average of 50-80% failed implementations) which left a bitter taste in the mouths of many executives. The promises of tremendous benefits of BI were enticing, but for many the bee stings often remained lodged far in the organisational backsides. Some of the reasons for the failures were cited to be the lack of academic research in the areas at the time, which limited the understanding of the implications of BI, Data & Analytics implementations. Many executives simply didn't know and/or understand the approach that would constitute success. There was little research into the right set of factors. Success was mostly underpinned by anecdotal reports from vendors about accomplishments of carefully selected initiatives, with many pointing at the set of technologies used. 

Huge costs, with, often, almost no benefits to show for them, made many organisations weary. Executives became very careful in committing funds towards BI, Data & Analytics projects if there was no case showing direct bottom line benefits (short-term tactical rewards).

Among many failures, there were organisations which succeeded beyond their wildest dreams. These companies reaped all the rewards of evidence-based factual decision making. So why that is where so many others have failed, some organisations still managed to succeed? There had to be a combination of factors and principles these companies embraced to ensure their success.

So, 5 years down the line the new digital technologies are revolutionising the playing field. Successful companies are dominating competitors with Big Data, Analytics, Machine & Deep Learning and other similar approaches promising a multitude of benefits. Multiple articles exist relating to how successful organisations are reinventing the way companies operate and compete, the way we see work and how the future is becoming more and more unpredictable, with quicker business cycles and turnaround times. The punch lines of these promotional pseudo white papers are [yet again] driving the need for executives not waste time and to grab onto the rails of a train leaving the station to success [but there is no guarantee].

Although some technologies are truly exciting, many of them are far from being new and often are just rebranded and/or updated versions of older products with minimal augmentation. Even though the choice of technology does play a part, many offered solutions [in the top quadrant] are of exceptional quality and have matured quite well over time as can be seen through many Gartner reports in the space. It can even be argued that they do not differentiate much and offer similar services when stripped to their core. 

After a lengthy introduction let's explore the importance of the below factors overall and see how their importance might have shifted with the onset of new digital technologies.

The top factors previously identified and combined into logical groupings from relatively scarce literature (refer to the paper here) were:

• Top Management Sponsorship, Commitment & Support
• Data Quality Management
• Team Work, Team Composition & Skills
• User Training & Support
• Clear, Strategically Aligned, Business Vision & Objectives
• Appropriate Technology and Infrastructure
• Project Management, Scope, Schedule & Planning
• User Involvement
• Change Management
• Development Methodology
• Easy to Use, User Oriented & Friendly Solution

Before we go further, I wanted to add a list of factors which were identified as important in the future at the time. These factors were Predictive Analytics, Business Maturity, Self Service BI and Big Data. Although I will not discuss them directly, it should become self evident how they have emerged in the past years.

Top Management Sponsorship, Commitment & Support

This is truly one of the top business factors which will make a substantial difference when it comes to the success of any project. If there is an executive sponsor who is committed and passionate about using data in solving business problems, they will support the project and lead it to success. This type of leader will most probably have a very good understanding of Data, Analytics & BI and at the same time is very good at navigating the executive suite and is clearly able to articulate the value of information related projects to other senior managers in the organisation (an official position of a CDO). A very important factor here is for this person to have a set of identified and well defined business questions he or she wants to solve by using the new data technologies.

To me, this factor seems timeless and no matter which technology arrives it will always be positioned at the top of the list. 

Data Quality Management

This is another big one. To explain the reason why, I will just say 'garbage in, garbage out'. If the quality of data is not checked, continuously tested and examined with the right loop back mechanisms and data lineage, people will stop trusting the information provided by the solution. They will find other ways to source it and use it 'rogue' style, which is a danger in its own right. Very often organisations do not really understand the total cost of having bad quality data. In fact, it costs organisations trillions of dollars annually and sometimes they don't even know it.

All employees need to understand the implications of poor data especially in the current setting. The onset of Big Data and IOT means that information needs to be analysed on the fly and there isn't always time to stop and fix it. It has to be planned in advance and anticipated. This means that even small nuances may cause information to be incorrectly interpreted if its quality is not up to scratch. (This is where some the costs come in).

I am hoping that in future (already happening in some organisations I would presume), through automation and AI, data will be able to fix itself. Until then I believe that Data Quality, as it stands now in many organisations, should be at the top of mind as it should have always been.

Team Work, Team Composition & Skills

I have worked for organisation and read in various articles, which utilise and promote techniques testing the compatibility (and therefore the quality and success) of a project team composition to improve collaboration and performance. A team which functions like a well-oiled machine, where members understand and complement each other from an xQ and skills perspective, will be more focused, productive and successful.

This facttor has become increasingly important in the past years and with new technologies like AI and automation. Many organisations are looking to hire individuals who are adaptable at integrating into existing teams and collaborate with existing team members while working towards a common goal without playing political games or being a burden.

Skills are important but can always be taught if the individual is open to learning and acquiring knowledge. In the future this will most probably become even more important as the learning cycles become shorter. In the 21st century the half-life to acquire new skills has changed from an average of 40 years to 5-10 years and will be getting even shorter as machines developed by humanity will be increasingly utilised in automating more and more tasks. People will need to shift roles and learn new skills.

User Training and Support

The users are, after all, the consumers of the solution or its output. They need to be trained and supported from the onset. They are the internal customers. In the present world, where customer experience and satisfaction is being continuously highlighted as one of the ultimate organisational goals, this factor takes one of the top spots. Supporting the clients is a change management strategy that enables the solution, familiarising the users with the toolset, making them more comfortable and effective in its use. The solution needs to be so familiar and easy in its use that its use becomes the path of least resistance. 

With the changing nature of work and self-realisation of employees' needs to continue gaining new skills, which minimises resistance to change, this factor should become easier to drive with time and the new generation of employees who will hopefully embrace change easier. Based on this I would therefore say that this factor is becoming less important with time. From my personal experience with organisational maturity when it comes to BI & Analytics it will still take time until users become self-sufficient and for now training and support is still as important as ever, although it might have shifted to a more digital and virtual form.

Clear, Strategically Aligned, Business Vision & Objectives

Structure often follows strategy which makes it imperative that the vision and strategy of a business is reflected in the solution. It is even more important to ensure that the solution build is flexible and supports the growing need for continuous change (CI/CD) and the DevOps approach. The vision and objectives should be short and to the point. They should also be articulated in a way where they can be quantified and measured. If you don't know where you are going and why, might as well not go anywhere as it will not matter anyway.

With such rapid change and advancements in technology, it becomes very difficult to strategize too far ahead as there is a need for more agility. The strategy review cycles are important but need to become shorter with the awareness of continuous digital transformation in mind. A great example of this is Amazon's six page memos which require executives to rethink strategy and its execution annually. One interesting question in particular is: how are you planning to use machine learning? This is a direct commitment to the use of data and information within the organisation which is driven by Amazon's second on command and is aligned to the company vision and strategy.

Appropriate Technology and Infrastructure

I still agree that this factor is important, but with the onset of Cloud and the flexibility it brings it becomes less important. Granted that the technology needs to be appropriate for the organisation and have the ability to integrate well and be supported by organisational resources. This factor is becoming less important with time as technology is becoming more commoditised using a pay-as-you-use and scale up and down at the click of a button. The barriers to access to infrastructure and technology required to deliver a BI, Analytics or AI solution are almost zero.

Project Management, Scope, Schedule & Planning

The importance of this factor is true for any implementation and is directly linked to the financial aspects of the project. This is especially true for these types of implementations as the scope creep can become a problem as people become more and more familiar with the advantages of having information at their fingertips. This factor has shifted to a more agile approach for many implementations but still stays very relevant due to the need to manage the financial aspects and delivery timelines in order to begin realising the benefits and return on investment sooner.

User Involvement

Involving the future users and introducing them to the journey early in the process will assist in their future buy-in and support of the solution. Their involvement gives the users a feeling of ownership. If they feel that they own the solution they will be more willing to participate and provide input into the solution. This is even more important when it comes to SMEs who will help in defining important aspects of the real need for information and problems experienced in the business which would need to be solved. When I refer to users being involved, I do not only mean this being a short engagement e.g. just to gather requirements or to clarify gaps. I mean users need to be involved in a continuous cycle. People who are continuously involved feel that they are valued and become promoters of the solution in the business. They help in quickly resolving any issues and close data gaps. This also allows people the opportunity to learn and stay at the forefront of technology which is becoming increasingly important.

Change Management (CM)

The softer factors are always extremely underestimated, undervalued and underfunded while they remain an important aspect of implementations involving new technologies and ideas. Change can be very difficult to many people and as it is often seen as a threat to the status quo. They can be very obstructive to new developments and slow them down through politics and inaction. In the digital age and the digital revolution with increasing speeds of developments in this area, the inability to embed change may mean the difference between success and failure of not only a project but the whole organisation.

Although in the current environment we are exposed to change more frequently and may embrace new digital ideas easier, as we might have heard about them in the media and feel like we understand them, the reverse is also true. An example of a technology like Robotic Process Automation or Automation as a whole frightens people as at the face of it, it simply means that people will lose their jobs to machines. This is not entirely true. Although there will be a shift in the job force in a direction we are not always sure about, and some jobs will become obsolete, new areas requiring expertise will open up and bring new opportunities.

To conclude this point, a well-managed and funded CM exercise will ensure that when new technologies are introduced, people will be more forthcoming in participating and helping to drive, and support them, rather than being obstructive. This is a very important factor and deserves to be at the top.

Development Methodology

It may seem quite interesting for some readers that this factor has been identified at the top. Surprisingly, it is directly proportional to the sustainability of the solution, ease of its support and ramp up speed in understanding its technical aspects by new joiners. It also affects the total cost of ownership of the solution as costs of bad development and testing can cost up to x10 more to fix in production.

A good methodology that works should be embraced, continuously practiced and improved. It shouldn't only exist as a set of documents in the organisational repository. I would think that nowadays a solution's development methodology should most probably align to a good, open standard rather than some proprietary offering. This way it should be easier to find and on-board resources.

The digitisation in this space, with the introduction of DevOps will inevitably shift this space towards AI and automate and streamline many of the components in this area, which should make this easier to manage but not less important. 

Easy to Use, User Oriented & Friendly Solution

This factor is quite relative as each user will judge the solution with their own subjective view. The solution will automatically be deemed easy to use, user oriented and friendly if factors like 'User Training and Support', 'User Involvement' and 'Change Management' are taken seriously and implemented correctly with a rightly allocated budget.

Looking at this factor from another perspective, users would be able to easily distinguish differences between various solutions based on usability and look and feel factors if presented with multiple choices, so this factor does, somewhat, play a role. Considering that user experience is becoming increasingly important, especially based on the fact that we are being introduced to a multitude of new applications daily, there is an expectation of a look and feel and ease of navigation to the point that it is intuitive. BI and Analytics applications have to be designed with specific design principals in mind to enhance the solution usability experience and ensure that it is seamless. It is not just about the creation of a dashboard or a graph anymore. Creating beautiful information is an art.

In conclusion...

It seems that although, in the age of the 4th Industrial Revolution, some of the factors influencing Data and Analytics solutions may have shifted over time many of them are becoming even more important than before. The high level analysis and overview of the factors shows that the technology related factors like 'Appropriate Technology and Infrastructure' are becoming less important whereas the business factors like 'Teamwork, Team Composition and Skills' are becoming more critical than ever.

Overall it would seem that the business oriented factors are gaining further ground on the technology related ones. This notion may stem from the fact that they are people related and presently are more difficult to replicate and automate.

Let me know in the comments what you thoughts are; what would you add or take away and why; are there any other consideration for the above factors. There is also a large list of factors in my original paper you can utilise for inspiration if you wish.

I hope that you enjoyed reading this lengthy but hopefully insightful article.