By Mark Lack
Extreme. Severe. These are two of the more common words used to describe the weather here in Texas and the southern plains.
This part of the country can go from drought like conditions one day to tornadoes and hail the next. The month of May itself has the highest average of tornadoes for the state with 277 (1995 – 2014). In fact this past Friday storms and tornadoes ripped through parts of Texas and Oklahoma unloading larger-than-baseball hail as they passed.
This kind of weather can be both dangerous and destructive. The Insurance Institute for Business and Home Safety ranked Texas No. 2* on its Top 10 list of states with the most property damage last year, with $23.7 billion between 2006 and 2013.
It’s one of the reasons Mueller Inc. is in business. As a manufacturer of steel buildings and metal roofs our products keep people and property a little safer. In fact, because our products can resist hail so well, customers can qualify for up to a 35% discount on Texas homeowner’s insurance.
But while our products can withstand the forces of Mother Nature, such extreme weather patterns can compound routine business challenges like forecasting, staffing, procurement, and more.
The good news is that data is raining down just as hard, if not harder than the stuff outside. At Mueller, we capture and analyze increasing amounts of information about everything from our products, transactions and revenues, to accidents, safety and customer satisfaction, from our more than 30 locations across Texas, New Mexico, Oklahoma and Louisiana.
But as data volumes have grown, it has become more and more challenging to make sense of it all – especially in a timely manner. Having used IBM SPSS for deep long term analytics, I wanted to try something more agile, dynamic and a little quicker to augment our business intelligence work. So earlier this year we turned to IBM Watson Analytics.
This cloud-based service has been a phenomenally fast way to analyze datasets. From the natural language querying to the visual results it produces, the tool has not only helped us get our arms around things like more accurate revenue forecasting, but it has opened up the power of analytics to a broader group of people and disciplines here.
You don’t need an MBA to use Watson Analytics. Once your datasets are created, the natural language querying is intuitive and helpful, automatically figuring out misspelled words, and even makes assumptions based on your queries that you might not have thought about. And with the new Watson Analytics Professional edition we’ll be able to start uploading and analyzing data from new areas of the business through its support of databases, business intelligence software and more.
When I show people Watson Analytics results and explain how I arrived at them through the querying, their reaction is invariably always the same: they want to ask it more questions. And that is one of the intangible byproducts of this service. The speed, accuracy and intuitiveness of the results inspire and prompt more queries from people – and not merely data scientists like myself – but business people, who consider variables that others may not.
And that’s a good thing. The more eyes to analyze growing data volumes means a more thoughtful and thorough business. And that’s what we’re trying to maintain at Mueller. While the weather can be unpredictable, we’re aggressively leveraging analytics to take the guess work out planning and decision making to be more efficient, strategic and a partner for our clients.