When it comes to BI integration, there’s just no end in sight to what you could integrate.
There’s always one more system, one more data store — and now, with Big Data and the potential for open Linked Data, not to mention Spare Data (look it up!) … I could go on and on, and so could the integration.
But IT’s time and resources are limited, so the question becomes: How do you decide what to integrate in a way that gives the business the most payoff? That question speaks to the heart of why integration should be a strategic issue, even though that’s not usually what happens.
What usually happens is integration is done on a project-by-project basis, but all you have to do is look at business intelligence and analytics projects for proof organizations could benefit from a more thoughtful approach to integration.
Time and time again, integration is cited as a major issue for BI projects. Just recently, SAS published a whitepaper, “Making Business Analytics Work: Lessons from Effective Analytics Users,” and sure enough, step number two was “Integrate across the organization,” with this telling stat:
… isolation of business analytics to a single department or function — commonly referred to as “siloing” — is more than three times as common (31 percent vs. 9 percent) among these less-effective users of analytics than it is in companies that are “very effective” in their analytics usage.
Recently, I interviewed John Lucker, a principal and Advanced Analytics and Modeling National practice leader for Deloitte Consulting. It’s his job to steer organizations toward better BI and analytics, so of course I asked him about this integration problem.
Lucker assured me there are a number of best practices that organizations can follow to prepare for BI success, including the integration work. Then he pulled this on me:
The first thing that I have seen is a lot of organizations really lose sight of the whole 80/20 rule around data hygiene and what data needs to be available to them. So I think what ends up happening is organizations never quite get done with what they want to achieve with BI and making data available to an end user. So I think focus on the 80/20 rule around data availability integration is important.
Oh no — not the dreaded Pareto Principle! That’s the first best practice for better BI?
Well hold on a minute there, Silver.
In this era of Big Data, Small Data, Social Data and data, data everywhere, it’s actually smart advice. Instead of trying to integrate some mass quantity of data, and assuming something useful will emerge, perhaps it’s time to sit down and really discuss what the goals of BI and analytics are. You might realize, for instance, that you need to add some external data or integrate different data sets when you think through how to support these goals, rather than the goals of a specific project or department.
The catch is, to follow it, IT will have to stop following the integration bouncing ball from project-to-project and take a more holistic view of what your organization needs to achieve, which, conveniently enough, ties in with Lucker’s next best-practice recommendation:
The other thing, as far as best practice goes, is that a lot of organizations don’t look at broadly across the organization to bring together the rich array of internal data and external data that they have and they don’t often look for ways to create synthetic information from the internal and external data … that gets at the key performance indicators.
I hadn’t heard the term “synthetic information,” so I asked him to clarify. He explained it’s creating new data variables or observations that don’t exist in a raw form from the raw data. That’s something you can only do if you’ve stepped back and taken a strategic approach to the data. Once you do that, you’ll be able to think about the last best practice — focusing on supporting the average BI user, rather than some hypothetical “power BI user.” He said:
My experience, from a best practice perspective, is to not presume you’re going to end up with a bunch of BI techies all over your company. If you expect any level of broad adoption of the use of BI and kind of information management tools and precepts inside an organization, there has to be a lot of thought around how this data is integrated and how it’s — to some degree — spoon fed for consumption.
A lot more time is spent on all of the technical nuances and delivery mechanisms and tool availability and not enough on, ‘How are we going to help our business users ask the really tough, vexing questions,’ and, ‘Is the tool really intuitively usable to people who don’t have degrees in computer science?’
It’s time to change that.