My Top 10 Blogs of 2014
2014 was a good year for big data, and for some of my big data blogs. This is my annual “My Top 10 Blogs” list. Choosing 10 forces me to exclude some interesting topics, such as my Big Data MBA series from my teaching stint at the University of San Francisco. Fortunately, my clients shared many real-world challenges so I had plenty of fuel for blogs. Here’s to 2015 being an even bigger year for Big Data!
#10 (Tied) – Dear John Letter
A good friend of mine who took over a company that provides a statistics software package inspired this blog. I have much admiration for my friend “John”, so I thought it only fair to share with him what challenges I thought he’d face when moving the software package forward, within a big data world. Hopefully this is useful advice for any company looking to transition or build software for this brave new big data world.
This blog really wasn’t about big data, but instead it challenged organizations to really understand where and how data, analytics and technology could be leveraged to provide business differentiation versus just business parity. I really wanted the big data conversation to focus on business differentiation instead of what version of Hadoop the organization used. Ugh!
Janet Yellen is the “Moneyball” queen of the Federal Reserve. Janet Yellen caught my attention as an example of a leader who understands that one needs multiple perspectives (metrics) in order to thoroughly understand the state of the business. Janet Yellen uses a dashboard of metrics to understand the job market and take decisive action. Although more data is better than less data, it’s perhaps even more important to employ a range of metrics and scores that help you understand the situation and make better decisions.
This blog gave me a chance to revisit my childhood game of Strat-o-matic baseball. This blog addresses the power of integrating “small” data – data captured via consumer comments, work orders, physician or mechanic notes, etc. – with the “big” data from your operational systems. There are many opportunities for organizations to ask that one extra question – to capture the “small data” – that, when combined with the organization’s big data, can yield more accurate and actionable insights. Organizations need a thoughtful process for ensuring the capture and central storage of this small data.
This blog was the direct result of watching our data science team struggle with a particular client’s data. The data science team was trying to explain to the client that the analytic process was fraught with failure, but that failures are progress because they teach you what doesn’t work. I happened to be reading an article about innovation at Pixar, and the connection between the two processes became clear – be patient, keep testing, keep fine-tuning and eventually success will happen.
This blog came very close to being #1 because it captures the analytic process that we use in every one of our big data and data science engagements. Our data science team is always looking to build analytic profiles at the individual entity level, whether that’s customers, products, employees, students, slot machines, wind turbines, etc. We like to designate these entities as the organization’s “strategic nouns,” those entities where superior knowledge and insights can help optimize key business processes and uncover new monetization opportunities. And this approach works every time. I love it!
This blog reflects that biggest problem that I see with organizations with respect to moving into the big data world: organizations are trying to apply old paradigms to new technologies and capabilities. It just doesn’t work. For example, I don’t know how many times I’ve seen organizations try to incrementally extend their BI and data warehouse capabilities by trying to integrate Hadoop and HDFS capabilities, and totally swing and miss. Then they blame it on the technology and not their approach. Ugh!
By the way, be sure to read the comment about “Don’t Think Business Functions; Think Business Initiatives” at the bottom of the blog. Interesting observation shared with my by one of my Big Data MBA students.
This blog took me nearly 12 months to write, and even now I’m not certain that I got it right. The Vin Diesel comment is a reference to his role as Xander Cage in the movie “XXX” where he admonishes the Prague police for not using their full firepower (heat seeking missile) in a standoff with a sniper. The way that a data scientist designs their data models is very different from how a data warehouse designer designs a data model. In order to take advantage of the raw processing power of MPP and Hadoop architectures, the data scientists want long flat tables that eliminate joins. And that’s what we’ve got – massive flat files against which the data scientists are running their analytics. They’re sort of ugly, but very effective from an analytics perspective.
This is one of my favorite blogs and a topic that I discuss frequently. I especially like how the blog addresses the advantages of a data lake to the data warehousing and BI teams. I’m a big fan of the data lake (even if I’m not a big fan of the name) and see more and more of our clients adopting the data lake not only to support their analytics environments, but more and more to free up expensive resources off of their data warehousing environments. The data lake is really a winner, and 2015 will be the year of the data lake (and I’ll be speaking about the Big Data MBA and the Data Lake at the February Strata conference…hint, hint)!
This is one of my more recent blogs, and it has gotten lots of favorable feedback. This blog took me a couple of months to get right. I really struggled to understand the key differences between how the traditional Business Intelligence process works (which I knew well from my Business Objects days), and how the data science process works. But after several engagements where I got a chance to work closely with our data science team, I think I finally nailed it!
This one is clearly my favorite. We turned the Big Data Business Model Maturity Index into a very cool, visually engaging animation. The animation turns the Big Data Business Model Maturity Index into a living, and hopefully thought-provoking story with engaging visuals. It’s a very creative presentation, and it actually makes me look smart. Heck, even my kids were impressed!!
So that’s it, my “Top 10 + 1” blogs. 2014 was a great year and I hope to have even more to share in 2015! Thank you for all your feedback and comments!