Essential IT Project No. 1: Streamline your company data
It's hard to resist the allure of big data. Gather enough data points, harness enough computing horsepower to crunch them, and you can predict what your company's customers will want before they even know they want it. You'll be a hero, and the business will own a license to print money. That's the promise, anyway.
The problem? Most organizations already have more data than they can handle, much of it inconsistently defined and captured in incompatible ways. So when decision-makers show up at meetings, they spend all of their time arguing about whose data is correct, not what the data is telling them to do, says Chris Stephenson, co-founder of Arryve Consulting.
The project you want to own is to simplify that data, make sense of it, and use it to propel the company forward.
Step one: Take the conflicting streams of data collected by different systems in your organization and consolidate them into a single database before business users ever get their mitts on it, advises Stephenson. To do that you'll need to work with business users to identify the important data points and arrive at common definitions.
"That's much easier said than done," he says. "But it will ensure that a company is managing to one version of the truth and allow multidepartment conversations to focus on the decisions the data is driving, not the data itself."
But even big data doesn't have to be that big. While you're waiting for that multi-million-dollar business intelligence initiative to pay dividends, you can employ "tactical BI" -- isolating the information that really matters to business leaders so they can make decisions more quickly, says Bill Brydges, managing director in MorganFranklin's Performance Improvement practice.
"Say Company A acquires Company B," says Brydges. "The CFO of Company A needs to look at the consolidated financials of both companies, and he doesn't have time to wait for IT to run it through its data warehouse. You need a tactical solution that can quickly pull data from multiple sources into a tool like Excel or SharePoint so your CFO can use it right away."
The key is to do it in a structured way so that results are consistent no matter what data is input or who does it, Brydges says.
"If you're going to create a spreadsheet where you mix the financials of Company A and Company B, you need to define the architecture so the next time someone asks for this, you don't start with a blank spreadsheet that produces different results and then have to create a third spreadsheet to reconcile the two," he adds.
That means business and IT need to work together to identify the bits of of data that drive results and figure out the best ways to mine them.
So, if this is such a great idea, why isn't everyone doing it? "Because it's hard to do," says Stephenson. "There are tons of companies stuck in the middle of $6 million BI projects. People are afraid it's too big to tackle. But our definition of 'big' is changing. In 10 years, everybody will be living in the world of what we consider Big Data today. There's no reason you can't start out by using smaller subsets of data to validate your ideas before investing too much time or money in a big initiative."