We attended a great analytics conference in Orlando this week. To start, let’s be clear that Florida as a destination was not as attractive as you might think – it was 40 degrees while we were in town!
The conference, TDWI Orlando, is part of a series of conferences held by the analytics industry group, The Data Warehousing Institute. TWDI has been around for nearly 20 years as the BI/analytics space has evolved from its meager roots to being a subject covered by magazines like The Economist and the Harvard Business Review in recent years. What we liked about this conference was the cross-industry focus – it wasn’t just healthcare, which is arguably behind the times with analytics. Chilmark Research, a healthcare IT research firm, wrote a glowing review of the conference last year, which prompted us to take a look at this years’ event.
We attended a variety of sessions, often dividing and conquering the available education tracks. Here’s a quick overview of what we found.
Traditional data warehousing and ETL processes are alive and well. This is the core of any analytics strategy, including big data. Even the groups who are focusing heavily in Hadoop continue to rely on their relational data warehouse for a significant portion of their analytics. Big data continues to lack definition – from debates over the validity of the three Vs to what defines “Big”.
When discussing predictive analytics, it was stressed and thoroughly debated that these new techniques are not a replacement for traditional dimensional data warehouses. If anything, new techniques can be used to augment available information on reports. In many cases, it’s prudent to feed data generated by predictive analytics and data mining exercises back into a data warehouse for reporting, trending and future evaluation of effectiveness.
Hadoop is here to stay, and is gaining traction within healthcare. Some research suggested that as laggards in technology adoption overall, we in healthcare are positioned to leapfrog many other industries who are weighed down by investment in traditional analytics tools. Qlik and Tableau are widely regarded by the industry as the best data visualization and data discovery tools out there.
Social media in business isn’t just for engaging consumers anymore. Sure, companies like Comcast and United have had their ups and downs with engaging their customers on social media, but companies are increasingly monitoring and mining feeds from Twitter, Facebook and even Pinterest to understand their markets better and how customers view their brand. Most social media have their public feeds, though serious social analytics requires the use of Gnip or DataSift.
People are still a huge part of any BI & Predictive Analytic program. Looking at the CRISP-DM (Cross Industry Standard Process for Data Mining), business understanding is at the top. Without a clear understanding of how the business operates, and what the needs are, and exercise could be thought of as a R&D venture with no direct impact on the business. Trust and Accountability are also huge factors to be considered. In many business sectors, new C-Level positions are being created with the purpose of assigning responsibility and accountability for these new ventures. After all, if no one is accountable, how can anyone trust the data.
We have heard a lot about analytics in the healthcare space as accountable care and population health management seem to have overtaken our world in the last couple of years. The TDWI conference was a great chance to escape from this world, and see what’s happening in analytics across industries.