Newsletters




20 Key Takeaways from Data Summit 2014


Bookmark and Share

Image courtesy of Shutterstock.

DBTA’s Data Summit 2014 in New York drew industry leaders from across all levels of the IT stack who shared insights on how emerging technologies and approaches are enabling business opportunities.

With three tracks of sessions, preconference workshops and a product showcase, the 3-day Data Summit conference allowed attendees to hear high level thought leadership as well as engage in one-on-one interactions about real-world challenges. As with any great conference, there were some thought-provoking takeaways that emerged.

Hear what experts really think about what's important in big data - including Radiant Advisors' John O'Brien, Google's Peter Boothe, VoltDB's Ryan Betts, Credit Suisse's Tassos Sarbanes, Penton's Raj Nair, Red Gate's Simon Elliston Ball, Rocket Software's Gregg Wilhoit, 1010data's Afshin Goodarzi, SAS's Anne Buff, IBM's Paula Wiles Sigmon, Embarcadero's Henry Olson, MemSQL's Eric Frenkiel, author Rick Smolan, Attunity's Tara Bruckert, Ntirety's Michael Corey, analyst and professor Elliot King, Search Technologies' Kamran Khan, and HP Autonomy's Dan Burke. 

In addition, many speakers have made their slide decks available at www.dbta.com/DataSummit/2014/presentations.aspx.  

20 Key Big Data Takeways that emerged from Data Summit 2014:

  1. It is becoming clear that the role of data scientist is too broad for any one person. It is probably a job for multiple people and what is also needed is insight from both the business side and the IT side, said several presenters
  2. Even if it were a job for only one person, there are simply not enough data scientists to go around. Data scientists often hold PhDs in mathematics and science, have been involved for a long time in academia and may have taught as professors; they like helping and mentoring others and are used to teaching and sharing, which can help others in an organization gain additional skills and develop into power users … Also: We need systems that are built to change vs. built to last John O’Brien, Radiant Advisors
  3. There is a difference between analytics and exploration. You perform analytics to make decisions. If you are not trying to make a decision, it is exploration. Also: In-memory analytics enables per-event analysis and per event analytics and decisions lead to revenue growth Ryan Betts, VoltDB
  4. People can do analytics because they have already done exploration; for exploration you need a big machine but only for a little while; data exploration and cloud are natural friends Peter Boothe, Google
  5. There is no point in storing data if we don’t analyze it Tassos Sarbanes, Credit Suisse
  6. If you want to get buy-in for your big data project, you have to evangelize it in your organization; do this by demonstrating the value with a project on a small scale. Also:  Hadoop is becoming the OS of big data Raj Nair, Penton
  7. We have evolved from SQL to NoSQL, to Not only SQL to now, “No, SQL” Simon Elliston Ball, Red Gate
  8. Data virtualization gives you the choice of any data consumer with any back end, without moving the data. The problem with moving data is currency Gregg Wilhoit, Rocket Software  
  9. Big data can be leveraged for predictive analytics to create products and promotions that cater to the individual versus marketing to the middle Afshin Goodarzi, 1010data
  10. With the opportunity of big data comes big data risk- there is a thin line between using your knowledge of the consumer for customer segmentation vs. discrimination. Companies need to pay attention to their ethical responsibility.  Also: If you do a big data project, you also need an analytics strategy - and make sure you are ready to turn insight into action Anne Buff, SAS
  11. There is a new C-level role emerging– the Chief Data Officer - to enable companies to look at data more strategically, and find ways to advance the business with data … IT and business teams feel they spend too much time looking for and  accessing data, but nobody feels they are spending too much time analyzing data Paula Wiles Sigmon, IBM
  12. Data is the lifeblood of many organizations and it is not a commodity. Your data is unique to your enterprise so it is important to manage it carefully. This creates special burden on data managers to really care for that asset. Also:  People think agile development is undisciplined but agile is a highly disciplined approach Henry Olson, Embarcadero
  13. Most real-time analytics deals with streams but the fast data also needs to be compared to additional data to add context and add value. In-memory is a component of a larger story Eric Frenkiel, MemSQL
  14. The opportunity provided by big data is like the enhanced vision a person experiences by opening their second eyeRick Smolan, author of “The Human Face of Big Data”
  15. Eighty percent of the work data scientists put into big data projects is data integration … Machine data is exploding – with 30% data growth per yearThere is big opportunity if you leverage all the data.  A well-known brewing company analyzes not just what sells in each store that carries its products but also the areas surrounding the stores to determine the timing and types of promotions and displays that will work best – does the surrounding community have a lot of Kentucky Derby spectators or are they pro wrestling fans? Tara Bruckert, Attunity
  16. What is needed is not big data but “thick data” that is rich with layers and context, said several presenters
  17. IT teams are devoting more than half their time to maintaining their infrastructure as it is, meaning that they don’t have time to make improvements for data accessibility to enable analysis and insight, or for improving data quality … Elliot King, Unisphere Research and Loyola University
  18. Server virtualization has become so easy – but you still need to read the documentation! And remember, there is no database too big to virtualizeMichael Corey, Ntirety, a division of HOSTING
  19. Search democratizes access to big data … Kamran Khan, Search Technologies
  20. The real opportunity of big data is insight and action - understanding what you have so you can take action and make better decisionsDan Burke, HP Autonomy


Sponsors