| | JUNE 20199E ERGYTech Review"What is Analytics anyway? Sounds like a buzzword to me. We've been doing data analysis for years. Now the marketers are calling it analytics so they can sell us something." I heard things like that. If I'm honest, I thought that myself early on. Thankfully, I wasn't evangelizing alone. There were others who share the passion and are doing their part to help steer our journey towards becoming a data driven company.Lesson Three: It's Better to be Mostly Right Than It is to be Completely WrongI think it was John Maynard Keyes who said "It's better to be roughly right than precisely wrong." What is the difference between analytics and data analysis? Is there a difference? I generally think of analytics as more sophisticated in technique, often involving far larger data sets than have typically been analyzed in the past, and generally more about the future. Although, historic data may be used, analytics is about doing something better in the future, or being able to predict things that haven't happened yet. Traditional analysis is more about explaining what happened. But that's just me. Analytics is certainly defined in the dictionary, but a few internet searches reveal several opinions but no real consensus on analysis vs. analytics. Here's the thing. I'm not sure it matters what the internet or dictionary has to say about analytics. The important thing is what does it mean at your company? While we want our analytics results to be precise, when it comes to what is and isn't analytics, it's more important to be consistent than to worry about being correct.As we've developed and rolled out our "Self Service Analytics Platform," we're leveraging the term "data consumers." We've identified three types of data consumption, reports, analysis, and advanced analytics. We expect 90 percent of our self-service platform users will be report consumers. They are going to run reports that have been developed and use them as they always have. The reports will just be easier to access. No more than 10 percent of our data consumers will be leveraging more advanced analysis tools. We're training those individuals in the use of the standard tools and techniques that are part of our platform. These tend to be the traditional "power users" who know how to pull data from various systems into their spreadsheets to answer business questions. Finally our 1 percent, the elite you might say, are our data scientists, or at least those who have been trained by our data scientists in more advanced techniques and tools. As a medium sized Midwest utility, we are fortunate to actually have a data scientist on staff, not in IT.So are our reports analysis based? Sure. Will we create re-usable reports that leverage advanced analytics? I expect we will. How do we draw the line between reports, analysis, and advanced analytics? I doubt that we will. If we get to the point where everyone who needs to, can leverage our analytics platform to improve our business we'll be mostly right, and that'll be good enough.Lesson Four: Blocking and Tackling still MatterAbout the same time we began our analytics evangelism efforts, IT worked with a local consultant to develop a Business Intelligence strategy and architecture. We certainly needed to architect a solution that could satisfy our business requirements, but it had to be sustainable. We needed to leverage our current environment and skillsets where we could, but also identify new skillsets and processes we needed going forward.Take data governance for example. No matter how advanced our analytics techniques and tools become, if they are applied to data that is inaccurate, inconsistently defined, or even misunderstood, the results and decisions based on them will suffer. The old adage still applies, "Garbage In, Garbage Out." Good data governance is the blocking and tackling of Analytics. Just as great quarterbacks fail without a good offensive line, so too will data scientists who are asked to analyze poor data.Next month I'll share some additional lessons concerning things like governance and architecture. You'll see the trend of not talking about the actual technology continue. Technology is easy. It's people that are complicated. Dave WebbAlthough, historic data may be used, analytics is about doing something better in the future, or being able to predict things that haven't happened yet
< Page 8 | Page 10 >