How Big Data and Predictive Analytics are changing the world for IBM and its clients

Downtown-Las-VegasLast week, Numrush was present at the annual “Information On Demand” event that was organized by IBM in collaboration with 300 of its business partners in Las Vegas. Attracted by the enticing slogan “Think Big, Deliver Big, WIN Big”, approximately 11,000 relations, staff members, clients and media representatives attended the event to learn about current developments and, more particularly, IBM’s vision of the future.

I mostly know IBM for its neat Thinkpads and from the time when large organizations looking for a high-grade mainframe or a similar large computer system had little choice but to opt for Big Blue’s products. Elephants like elephants, which meant IBM did good business with other blue chips.

The four pillars of the new IBM

IBM has not exactly been on fire on the stock market lately. A number of analysts are frantically debating whether the company’s stock should be considered a value stock or a growth stock. The fund pays out excellent dividend and has a sound cash flow, but it seems to lack the truly golden growth opportunities to see it as a growth stock and to value the company as such.

In the many people I spoke to and presentations that I attended in Las Vegas, I noticed that IBM is working incredibly hard to regain its position as precisely that growth stock. After the sales of its PC division and acquisition of Price Water Coopers’ consultancy division in 2002, the company seems all set for a new transformation. Fortunately, IBM has mastered the skill of revolutionizing its business model like no other company. The same applies to integrating acquired companies, but more on that later.

The new IBM is based on four pillars: mobile, social, Big Data and data analytics. At first sight, these did not seem particularly innovative or exciting to me, but after a couple of days it turned out they represent much more than fashionable buzz words.

First of all, there’s mobile. This is something I actually regarded as the least exciting pillar. Needless to say, it won’t be long before everything is accessible and available through our mobile phones. This is already par for the course for consumers, and seeing as many consumers are employed by organizations, they will expect the same level of accessibility and availability in their work. In this connection, it pretty much seems like it is mainly large organizations that have a problem with this development.

Social as a data source

For IBM, social comprises two areas; inside and outside the organization. Social outside of the organization particularly involves social media analytics and other sources that import data for IBM. Needless to say, IBM regards social as a major data source to which it can apply its predictive analytics software to monitor sentiment, among other things.

IBM also offers a range of software solutions for social media within organizations. The company’s sale sector relies heavily on some really complex softwares built on SAP. In fact, the company uses shipping solutions based on SAP. Under the flag of Connections, IBM supplies organizations with Enterprise social software, which comprises wikis, blogs, communities, files, forums, profiles and various other solutions that ensure employees are able to find and share the right information within the organization and communicate it. This intel is also recorded, so that it does not get lost when employees leave employment. You can compare Connection with a content management system that integrates LinkedIn, Twitter, Wikipedia, Dropbox, WordPress and so on. In short, employees can use all the social media they use on the web, but as an internal variant. I have never used the system myself, so I’m unable to tell whether it is nice to use (please leave your experiences in the comments). I also don’t know how it will develop going forward. As an employee, you may not want to register and share everything twice, so there is a risk that people won’t use the system with great frequency. Fortunately, various plug-ins are already available that (partly) solve this problem. For instance, there are LinkedIn widgets that display information gathered from LinkedIn in the IBM environment. And if it works, organizations can build wonderful collections of data on which they can run various kinds of analyses.

Probably the two most important pillars for IBM in the coming years will be Big Data and data analytics. As indicated during the conference, these go hand in hand: without Big Data there is no analytics, and without analytics, Big Data has little use. Contrary to what you might expect, having more data at one’s disposal actually simplifies analysis and makes it faster rather than slower. Putting together the last pieces always takes place faster than solving the first part of the puzzle.

The cloud is crucial

In the area of Big Data and data analytics, the cloud plays a fundamental role. IBM has already developed solutions for this, but the company did not seem to be a serious player in this arena or able to enter new markets until it acquired SoftLayer. This acquisition has allowed IBM to haul in a new clientele of start-ups and SMEs that traditionally didn’t form part of its target group. These customers will increasingly use all the products and services in IBM’s portfolio. The acquisition also allows IBM to attract new customers in this segment. More importantly, however, SoftLayer seems to function like a catalyst within the IBM organization when it comes to marketing new and current services. It has opened up a wealth of opportunities for IBM to offer their client base new and existing services via the cloud in a flexible way.

Besides Softlayer, various other acquisitions seem pivotal in the area of predictive analytics. In this context, SPSS (reading the name may cause a lot of readers to break into a sweat in their memories of statistics assignments during their study) and Cognos BI are frequently mentioned. This also brings a bit of Dutch influence into the bargain. After all, in 2003 SPSS took over Data Distilleries, a Dutch data analytics solutions provider. Thanks to the technologies from these acquisitions, an increasing number of IBM clients are succeeding in transforming their operational activities using analytics. This can be anything from police departments that use the technology for crime prevention purposes (which did give way to a lot of Minority Report questions during the event) to the more run-of-the-mill examples like the ability to predict stock levels for shops selling jogging items in the run-up to Christmas well over a year in advance. Or the ability to predict anything from near-hits by asteroids or predicting your opponent’s serve using IBM SlamTracker, as Serena Williams discussed during one of the keynotes.

One of the most exciting announcements in this connection was Project Neo. This project has not seen completion yet; at the time of the convention it had only entered its beta phase. However, IBM expects so much from this project that it already wanted to reveal a couple of examples. And rightly so, as it’s a beautiful illustration of how technology can provide any organization with its own virtual data analyst. Project Neo allows you to use natural language to request data sets and create graphic representations after making a selection. Subsequently, you can click to zoom in on details or add new parameters using voice control. During one of the real-life cases, a demo was performed involving a movie director who was in search of ideas for the next blockbuster. First the director asked which movie genre typically produced the best results. After selecting a data set from IBM, the answer was presented in the shape of various blocks. The director was then able to select other views in real-time. After asking the question “What costs are involved in producing a motion picture”, the investment was made visible, followed by a visualization of the net turnover per movie genre. Excel on steroids — without the need to be an expert in vlookups and hlookups, you can create a neat visualization at the drop of a hat. According to IBM’s researchers, this will make presenting information much more dynamic. They even used the term storytelling.

Project Neo is currently only available as a beta for the happy few, but will see its market introduction next year in the shape of a cloud service. Incidentally, Project Neo does not use Watson for natural language, but features an engine of its own that has to be trained for each client/domain using a proprietary dictionary etc. So it’s not like everything takes place automatically just yet.

Watson on the smartphone?

Needless to say, there was also a lot of hardcore technical news. Hadoop, structured and unstructured data, in-memory technology, NoSQL, NewSQL and BLU Acceleration being just a few examples. I am not really much in the know about these things, but it was interesting to notice that it is precisely IBM’s year-long hardware experience that allows it to process an ever greater amount of data at continuously increasing speed. This allows IBM’s technicians and researchers to develop hardware that has been optimized for its own software and achieve considerable advances in terms of performance.

So what about Watson? This supercomputer is still being further developed and continues to produce striking results for IBM. What remains unique about Watson, is that it has been programmed in such a way that it is not steered in advance, but starts to look for patterns entirely of its own accord. During the event, IBM also said it expected that in the future, Watson will not only use existing data sensors, but all sorts of sensor data. This means Watson will finally be given ‘eyes and ears’.

Unfortunately, Watson is still a too sizeable system to be integrated in mobile phones, which means it’s still not possible to ask it questions on the go. Unless, of course, Watson is made available as-a-service in the shape of a cloud service. Rumors to this effect have been surfacing, but have been denied just as quickly by IBM. However, this does not rule out the possibility of that happening.