On Tuesday I attended the IBM Business Analytics & Optimization Roadshow in Conshohocken presented by Micro Strategies.
Opening the event was Dr. Martin Fleming, Chief Economist and Vice President, Business Performance Services.
He spoke about transformation in a global economy and identified five periods in our history and how each has three distinct phases: install, crisis, and maturity. The periods were marked by the Industrial Revolution, Steam and Rail, Steel and Electric, Oil and Automobile, and Information and Telecommunications.
He further noted that the credit crisis was the crisis for the Information period and once we move out of the crisis we will see the space mature.
Dr. Fleming then addressed various global regions stating that Asian economies are on a path to stronger growth, while the Euro zone has had six straight quarters of contraction going into 7th with 50% plus youth unemployment.
For the US, he noted that the debt levels are coming down, consumers continue to spend despite the tax increase, but business remains more conservative.
He concluded by saying we fixed the short and medium finance issues, and we are now poised to take advantage of connecting digital with physical in moving ahead the economy.
We are also seeing as the economy recovers, the replacement of old infrastructure with new infrastructure that incorporates sensors and technology unlike the old. During the Q & A he explained that this integration is what will make the world a “smarter planet”.
This tie-in made sense, but I wish he would have noted it from the beginning.
Scott Sampson, Director, Information Management, then presented, “Harness the Power of Big Data”.
He began by noting that information management is software and then explained ”Smarter Planet” as being IBM’s strategy/approach to business.
He stated that we are living in a world where empowered consumers have great expectations. It is a hyper-connected world. We are seeing increased complexity, the emergence of big data analytics, and everyone is being forced to do more with less.
An IBM CEO study indicated that technology was 6th most concerning issue in 2004, but rose to the top spot in the latest survey from 2012.
He also stated that data rates will grow 40% year to year, while operational expenses are getting squeezed, and capital has remained steady.
Sampson then delved into the questions, “What is big data?” and “How do I deal with this in an economical manner and gain insight?”
“Trying to do everything in one solution doesn’t work well”, he said using the example of a houseboat being neither a great house or boat.
An MIT study shows organizations that see analytics as a competitive advantage is growing rapidly.
There are four core types of data: transactional and application data; machine data; social data; and enterprise content.
Why is big data tough?
- There are low numbers of data scientists,
- Complexity of solutions (economy of getting it done)
- Big data underpinned by data integration
Must have a solid business case to engage a big data project.
How does IBM look at data:
- Volume, amount OD data and users
- Velocity, data in motion
- Variety, structured, unstructured, text, multimedia
- Veracity, data in doubt, is data credible
Big data platform
- Open source leveraged and extended
- Integrate back end data
- Platform has to grow
- Industry standards emerging
Don’t think of a big data platform as a single product, but instead a set of capabilities
Analytics is mostly reactive (looking back) 80% vs. predictive 20% (looking forward). Most don’t begin looking forward until they have mastered looking back.
Scott Parker, WW Business Solutions Consultant, presented “Binoculars into Big Data”.
IBM purchased the search engine Vivisimo in 2012 and renamed it Infosphere Data Explorer. Early on I had used Vivisimo very regularly and really liked the way it grouped search results, so hearing it was a part of IBM’s solutions intrigued me.
Parker noted that there were typically three main discussions when engaging a Big Data project: Business discussion, Technical discussion, and another Business discussion where goal is to quantify a solution.
Getting started is all about visibility, you want to see what you have, to gain insight.
He went on to identify 5 key use cases:
- Big data exploration
- Enhanced 360 view
- Security/intelligence extension
- Operations analysis
- Data warehouse augmentation
He emphasized the need to take a phased approach. You want to start by first understanding your data and then leverage it.
Infosphere Data Explorer indexes data like the search engine it is. Security, scalability and relevancy are key attributes of the product.
You also have to consider navigation and discovery, what you have and how you get to it.
Some examples of data types within the enterprise you can connect to:
- Relational data
- File systems
- Content management
- Supply chain
- RSS feeds
- Custom sources
The solution also has social features to allow users to like, tag, and share. While this may not on the surface seem important, consider the massive task of meta-tagging all of your corporate data and content. Now consider each time any employee interacts with a piece of content, image asset, data, etc., they can add context and assign value, which will be indexed and used in future search results.
The 360 view is basically a CRM giving info about customer interactions.
The search function is fed by text relationships and meta data to increase relevancy.
It’s about getting the right info in front of end user regardless of where your data and content is stored, instead of trying to move all of your data and content into one centralized location in order to then search it.
The solutions can act as an Intranet or customer portal because of the use of rights management.
How do you pull all this data together? You have to be ubiquitous, because your end users don’t want to be logging in and out of different systems, so the single portal needs connectors to all of the different sources.
A pre-configured Connector library includes SAP, Share Point, email, and many other enterprise standards.
A users portal, or dashboard, knows the user by role/position, so it is configured based on that, plus it knows your activity history, so it also leverages that to put the most relevant analytics and content front and center.
Identify challenges and get insights into hands of people who can make them actionable (sales, service, marketing).
Discover and navigate then analyze and visualize.
Sahil Kedar, Technical Consultant, presented, “Find the Future in Hidden Data”.
Kedar began by noting that analytics provides $10.66 return on each dollar spent.
75 years ago it was easy to know your customer, think corner market, hyper-local, face-to-face interactions on a regular and consistent basis.
Today you have to determine which customers are likely to leave or cost you money?
He spoke of the need to examine customers networks and connections to determine related customers, because you want to ensure everyone in network remains a customer and hopefully an advocate of your product/service/company.
Location based data is next big thing, he stated, and then noted it has been the next big thing for several years, so he is really hoping it happens soon.
He then addressed the custom IBM PureData Platform and provided a more technical explanation of how the box works.
My last note from his presentation was the fact that data is not disposable, it may not yield result today, but it could tomorrow. We have the storage capacity, processing speeds, and search capabilities to handle large volumes of data, so there is no need to discard any of it.
While this was clearly an opportunity for IBM to showcase their solutions around Big Data and Analytics, there are many providers out there.
Having said that, I am very interested in learning more about the InfoSphere Data Explorer and how we might be able to leverage it within our company.
My next step is to reach out to internal stakeholders and begin having discussions around painpoints and unmet needs. With that we can begin to see how these solutions could be employed and then clearly, do we have resource, budget, and approval, to implement.