Analytics, also widely known as Business Intelligence, is the collective name for the software applications that turn the data obtained from enterprise operations into an intelligent, actionable picture of what is happening within the business. It covers historic analyses, reporting on (potentially) everything that happened in the past; the ability to drill into that historic data to find the answer to “why” questions; and predictive analyses. The results are usually organised in line with the structure of the business, and by time (the ‘reporting period’).
From that basic level of capability, increased sophistication and complexity has added more functionality, until today’s analytic suites are extremely powerful and multi-functional.
Broad categories of Analytics include:
- Reporting – the most basic form of Analytics, a simple structuring of the data into a hard copy record;
- Dashboards and scorecards – producing a high level summary of key facts (Key Performance Indicators), used to make the essentials readily accessible;
- OLAP (Online Analytic Processing) – the forming of the data into multidimensional cubes, structured by key dimensions (time, place, organisational unit, etc.) to ‘slice and dice’ the data (examine it across different dimensions), in order to find things of interest;
- Advanced Analytics – the application of algorithms that discover more sophisticated patterns in the data than are apparent from just ordering within dimensions. Usually, these techniques work by identifying the strength of a link between various fields, and using that to classify them in some way. Advanced Analytics can be applied to historic data (which helps us to understand how we arrived at where we are now); or used to predict likely future outcomes;
- Visualisation – now emerging as a key capability, where the complex numeric tabular data is presented as various kinds of more visual representations, from simple RAG (Red, Amber, Green) status indicators, to metaphors such as speedometers with marked targets. This is based on the idea that you can absorb complicated associations visually, more readily than you can from a tabular numeric format.
Many specialist areas within business have their own tailored Analytics solutions, so there are strong vertical and horizontal offerings covering the needs of marketers, finance, logistics etc., and there are solutions tailored to Telecoms, Financial Services, FMCG etc. Another way that these technologies are bundled to support specific needs is as a Corporate Performance Management Suite, which deploys functions such as reporting, OLAP, Predictive Analytics and Visualisation, for a particular target audience (such as the executive level management that wants to understand the big picture), and which is more focused on financial outcomes than operational detail. There are also Analytics solutions that deal with various data niches, such as machine data or geographic data. If there is data there is likely to be an associated form of Analytics.
We live in a world with ever-expanding amounts of data. Not only is data now growing in volume, it is also very diverse, so we are all struggling to cope. We are in danger of being swamped by the data and missing the message. The role of Analytics is to filter that mass of data and turn it into meaningful, actionable insights. For any operation to run effectively, those who direct it need to make effective decisions and, as the world becomes a more global, fast-paced place, the ability of everyone to make consistently good, timely decisions is being stretched to the full. Analytics assists by presenting just the most important facts, in as accessible a fashion, and in as timely a manner, as is possible.
Analytics is the icing on top of a complex technology stack; and there are many complementary technologies involved in a successful Analytics deployment:
Anyone who manages any process or organisation with any degree of complexity should be interested in having Analytics to support their decision-making activities. An analogy is that of driving a car; would you drive a car if you could not see the road, and had to rely only on what you thought was going on around you? [Actually, that is just what we do; only about 10% of our perception comes from our eyes and the brain fills in our world-view from its other senses and experience, which is a major cause of RTAs]. This would be like running a business without all the facts being available to guide you. Would you drive a car with only the rear view mirror to guide you? This would be like running an enterprise with only historic analysis to guide you.
So every enterprise should be interested in the effective exploitation of Analytics capabilities; and, at all levels, people should seek out the level of support that suits their needs and which they can afford. Because of this widespread demand, there is an Analytics market with offerings of all levels of sophistication, offered at a wide variety of price points. This means that there should be something to meet everyone’s needs.
From the simple spreadsheet to the most complex of statistical analytics, all forms of analytics have to contend with volume, and Big Data is a key theme in many analytics solutions.
Most people tend to think that Big Data only refers to the scale of what is being processed, but of far more importance than just the scale is the diversity of the data that is now being processed. Not just transactional data, from the enterprise interface to its customers and suppliers, is being exploited; but also all of the low level operational data from the mechanisms used to provide the operational solutions (such as network data, sensor data from machinery, and so forth) in addition to unstructured data such as social media, the analysis of which is a major part of the big data phenomenon.
This is then having a profound impact on the nature of the silos within a business. If you look at a Telecommunications company, everything was once split between BSS (Business Support Services – largely to do with Finance, Marketing and Customer Care), and OSS (Operational Support Services), including the network and all of the low level technical detail. The two worlds coexisted but the touch points were limited to key things such as the OSS passing call details to the BSS in order to produce a bill. However, today we see those barriers being shattered, as Analytics can be applied to the lowest levels of the operational data (such as sensing a failure in a network component), and then pass that data in real time to the customer care systems in order to identify the customers who might be impacted, and then, according to their value, decide on the appropriate proactive remediation steps necessary on order to manage the situation.
Another key trend is the move of Analytics from a niche technical specialist domain to the general business domain. Intuitive point-and-click interfaces are now available, which abstract the problem definition and solution components so as to require minimal technical know-how from the user. These can produce solutions of quite dazzling technical sophistication in minutes, for non-specialists.
A further important trend emerging as a differentiator is rich visualisation. As the scale and diversity of data accelerates, it becomes ever more difficult to separate what is important from background noise. To help time-pressed decision makers to see critical trends and associations, results are increasingly being presented as graphical representations of the underpinning tabular data.
The big vendors – IBM, SAP, Oracle and Microsoft – all have analytics offerings that reflect on the importance of this category of solution. All of these are highly credible and, for many, they will provide all that is required.
For vendors such as IBM, analytics is core to their forward business strategy and has been identified as one of the main pillars for growth. A supplier like IBM can be seen as being as innovative and disruptive as any of the niche players in areas it sees as key with, for instance, its artificial intelligence solutions based on Watson.
It used to be that the major vendors’ offerings were lacking in features, and the specialists were seen as being the leaders in innovation, because they could bring things to market quicker. Now, however, the big vendors are as likely to be working on the leading edge of this technology as anyone.
However, such is the diversity of the market place, and the disruptive impact of Open Source and Big Data technologies, that any attempt made by the big vendors to corner the market is hamstrung by the dynamics and economics of covering so vast an opportunity. So, the market is full of established and emerging niche or specialist players, nearly all of which can offer things of great interest, and of great value to significant elements of the market. Thus, you cannot write off the likes of MicroStrategy, SAS, Splunk, Pentaho, Tableau etc.; and there are the new entrants, such as Guavus, Vitria, Nominum, Accountagility, Board and many more, that pose a real threat to elements of the existing market.
Although the big players may dominate in terms of volume there will always be areas that they just cannot cover, and if that is an area that offers competitive advantage, a niche supplier will find an audience for its products. Differentiation is difficult because vendors adopt very different philosophies. For instance, Splunk, which is probably the market leader in analysing the masses of machine data available now that sensors are embedded in just about every aspect of a workflow, favours an approach where the analysis is only loosely structured, and it looks for previously unseen patterns. A vendor such as Nominum, however, which also analyses machine data, favours a very deterministic, structured approach, looking for specific patterns leading to a specific outcome. Both are excellent products, so that selecting between them needs an understanding of what is important to your business and why. Increasingly these decisions are not technology choices but should be made by understanding the alignment of specific products with your particular business strategy.