Gartner describes how crucial it is for many companies to exploit internal data profitably over the course of a financial year. So in 2015 already, business intelligence (BI) and big data analytics ranked highly in the new trends. This development is hardly surprising if one considers that 90 percent of all data generated around the world is no more than two to three years old! It follows, therefore, that a mere 10 percent of all data was acquired earlier than 2012. Controlling this data takes a lot of time and energy, and is also a costly affair.
Now it is high time to find more sophisticated ways to handle it. Get started with big data analytics! Install the necessary software, kick-start your own big data platform and begin analyzing the honeypot of information at your fingertips. Storage has never been this affordable, and a hybrid solution will let you get going at the flick of a wrist. Your data, databases and protocol files possess immense value. Read on to find out more about big data analytics.
An article by Remco Toele, Solution Advisor Enterprise Software
The age of big data, big data analytics, BIG, and the Internet of Things (IoT)
Big data (BD) is the talk of the town right now. Essentially, it enables the investigation of historical data and its analysis based on certain patterns. What’s more, big data analytics lets users peer back in time and even into the future. Companies can set themselves apart from the competition by making intelligent use of their own big data analytics platform.
In what ways can companies benefit from their data? By now, plenty of organizations have recognized the immense value of smart gear belonging to the Internet of Things. These devices put them in a position to improve and accelerate their production processes and even to predict whether a component will fail. Good examples include machines and systems used within the automotive industry and elsewhere.
Here’s another concrete example: Big data analytics can take information obtained from electricity and heat sensors to control closing times in multi-level buildings. In this way, cleaners can be deployed more precisely to suit the work times in each individual department.
Real-time information obtained from the big data platform helps early identification of emerging trends. This way, organizations can respond faster to changes on the market. Real-time big data analytics, for instance analyses of database logs, allow companies to tap into fresh insight concerning their own security and any potential attempts at fraud. This way, data loss and other impairments can be detected at an early stage.
A brief look at big data
Just five years ago, big data was seen as the Holy Grail and seemed on the verge of becoming a veritable hype. Plenty of companies jumped on the bandwagon, and software vendors quickly developed new, extremely good (and expensive) platforms. But a big data platform alone is not enough. Many companies recognized this fact, despite their tentative success. After all, BD can deliver more than just one successful project.
Big data in practice
Gradually it is becoming clear that some companies have missed the boat. 2015 was the first year in which the value added and success of big data analytics were plainly evident for everyone to see. The market can be divided into three different types: New Players, Transforming Companies and Conservative Companies.
What sets New Players apart from the other types is that they deploy a relatively small workforce to service a hefty chunk of the market. Good examples here include companies providing online streaming or music services, for instance vimeo or spotify. Not only do they manage sensitive customer data, they also maintain relatively lively contacts. This is remarkable and only possible through the use of intelligent software that builds on a big data real-time platform to deliver analytic insights. Here, the marketing department can use extremely targeted and personalized offers or suggestions to appeal to customers and – provided it is done right – even surprise them. The key is to contact the customer using the communication channel that they favor, for instance email, social media or by post. Restricting customer communication to one or two channels generates a particularly high response rate and customer satisfaction. New companies will find this approach extremely successful when combined with an innovative product or solution.
Distinct features of these companies are that many of them will have been around for a long time, that they maintain a large number of branches, and that their workforces may run to the hundreds or even thousands. Telecommunications corporations, TV groups or publishing houses are among the organizations that need to transform continuously. These days consumers communicate via WhatsApp as much as they do by telephone. Some viewers are no longer satisfied with a set TV program and want to decide what to watch, when it suits them. Companies like Vodafone or Conde Nast need to be versatile here.
Many of these companies will rely more on traditional sales and customer loyalty. Often they will be really quite large enterprises with big department stores in the city centers. In recent years, some of them have been the subject of news coverage due to the (apparently) perilous predicament in which they find themselves. The question is whether they can even reverse the situation on time. The automotive industry is another example: will it continue to be the only sector building cars in the future? Nobody would have thought just a few years ago that Google and Apple would also develop software for use in vehicles. Now the established groups are called on to adapt.
Gaining a foothold in big data analytics
Anyone keen to find out how companies can acquire added value from big data analytics can simply explore the issue at home. Quite often it seems as if more data – so-called ‘dark data’ – is available than is actually used. But this data can still be useful, for instance to optimize processes and to achieve ‘quick wins’ and improvements. Success like this can then provide motivation to improve and optimize even more business processes.
There is no shortage of data sources – for instance databases, logging, IoT and social media – that can be analyzed and linked. It is easy to do with the right big data platform, and users will quickly acquire new insight or obtain interesting pattern recognition. They can choose between proprietary software or hardware and open source solutions, although combinations of proprietary and open source are now increasingly commonplace.
Innovative search technologies like those offered by IBM, Apache, Elastic and Splunk today enable very simple execution of even the most complex queries. The hits and results can then be processed using popular analytic programs like Tableau, Kibana (open source), Cognos and Oracle, or BI solutions by SAS, SAP and Qlik. What makes these analytic solutions strong is that they visualize data to facilitate comprehensibility and accessibility. The results help companies keep their products or services innovative and to develop smart devices that are currently in demand, or will be soon.
Big data analytics licenses and subscriptions
A number of providers with very different requirements, licenses and subscriptions are involved in many big data analytics environments. The terms that these software vendors offer may be influenced by the infrastructure and user structures alike. It is important both before and during the project to keep a keen eye on the costs, possible risks, license management and the functionalities that the various options entail. This helps ensure that big data analytics projects proceed within a calculable and predictable framework.
A large number of companies are already working flat out to test new and innovative software solutions. Naturally, test projects that deliver particularly good results with Hadoop or another analytic (search) tool are taken to be especially meaningful. Brought together with existing data sources, the projects become part of the production environment. In many cases, though, departments and administrations are not always clear about how to use big data analytics software. After all, in many cases the solutions were created on a laptop by a development or web team working at a remote location. Therefore, not all Hadoop distributions are adequately protected against unwanted access when used straight out-of-the-box. This may expose companies to substantial, even financial risks. License or subscription fees can also add up, also with potential consequences for the future.
Strong support is available for the most familiar solutions like Apache Solr, Spark, Elastic and Splunk. Users are asked only to sign up for subscriptions, which – like licenses – come with a variety of price and support levels. There is also a number of public support forums, some of them providing 24/7 assistance and fast response times. The prices and levels on offer from the most popular providers also differ. It is therefore advisable to take a good look at the specific properties of the different big data analytics products at an early stage. This way the users can make quick and nevertheless sustainable decisions when the time comes.
(Big) data-driven organizations
As we have seen, complete administration of information systems can be extremely time-consuming and costly. But the investment may well pay off. After all, big data analytics can yield insight of immense value to decisions concerning information processing and management. For companies, this development means taking the first step on the way to becoming data-driven organizations. After all, decisions that are based on actual data will generate larger profits or returns, as the companies will acquire greater understanding of their own organizations, customers or products.
Anyone looking to test big data analytics should first be clear about its actual purpose. This might be enterprise-wide, departmental or within the framework of a specific project.
Examples of data-driven projects:
- Enhanced search functions inside and outside the organization / understanding and improving patterns
- Log(file) analysis and monitoring security and fraud
- E-commerce (consumer habits / marketing)
- Publication of documents (process / life cycle)
- Search for new HR and optimization features
The future of big data analytics
A fresh mindset, a different approach to our data – our most valuable capital – will be necessary in future. Today’s digital environment and the way in which we access and save data are based on three to five-year-old technologies and philosophies. This is not particularly surprising if one considers that the average service life of our hardware and servers is approximately three years, and in some segments can rise to even five years.
It is imperative that besides understanding the current accumulation in data and management, EU companies in particular need to keep a close eye on their security and its relevance in terms of the EU Data Breach Notification Rule. The cloud strategy is another important factor, of course. Here, data can be stored locally, migrated to the cloud or kept as part of a hybrid solution.
Big data analytics in a nutshell
Increasingly, companies looking to be fit for the future will need the agility to respond to changing and new markets – preferably in real time.
Startups will find this substantially easier than the majority of small to medium-sized businesses and enterprises. The question of which kind of real time should be applied on the way to becoming data-driven companies will be crucial: Will data from the past hour, from yesterday or last week exert the greatest influence on (future) business?
Using internal or external data to identify patterns of thought or outcomes in the past (BI) or the future (Big Data Analytics) takes foresight, courage and a clear vision.
In recent months we have observed a number of large companies experiencing difficulties when faced with new circumstances, and in some cases even going to the wall.
But there are winners, of course, besides their less fortunate counterparts. The successful ones tend to be quick on their feet, perhaps because they recognize early on when there are signs of changes in consumer behavior. Put succinctly: An enterprise with the capacity to analyze data quickly and to provide management with the right results as a basis for strategic changes in course will fare best in terms of growth, revenue and profits.