In 2015, Gartner reported that only 15% of analytics projects ever make it out of the pilot state to enterprise deployment. That’s a lot of wasted time, effort, and money. It doesn’t have to be that way. Over the course of my career, I’ve spoken with many experts and C-suite people about why analytics initiatives fail. There are thousands of stated reasons, but, really, it all comes down to one root cause: the analytics initiative was treated as a project, as something the organization was, “doing,” not something the organization was, “becoming.”
Businesses have been playing with some sort of analytics capabilities for over 25 years now. Over that time, tool suites and methodologies have come and gone. Innumerable analytics theories have been offered and rejected. But one thing has stayed the same. To truly be successful with an analytics initiative, analytics must be imbedded, at the operational level, throughout the enterprise. You must become an analytics organization.
Indeed, analytics’ value to the business can only be realized when analytics tools are running within existing production systems, and embedded in existing, or improved, business processes. These processes must function within your normal business operations and within the business’s strategic framework and support the goals of the business. It sounds simple to say—and we’ve been saying it for 25 years—but it’s absolutely true. There can be no successful analytics initiative that is not intricately imbedded—not just tied to—but imbedded within the DNA of your company. It just isn’t possible.
This “Analytics Ops” philosophy rests at the intersection of your company’s data science, data engineering, and process engineering efforts. It’s also not a one-size-fit’s all methodology. Instead, it’s a mind-set. It’s not about what tools to use and what principles and practices to follow. It’s about deploying analytics that fit your particular business needs, and using a deployment process that enables you to thing big, start smart, and scale fast.
Analytics Ops, and embedding analytics throughout the enterprise, requires a holistic approach to the project. It starts with determining the scope of your project. Understand what problems you have, and which of those are most cost-effective to solve with the pilot. These are within your initial scope. The project can scale by adding to the scope, but NOT before demonstrating success in solving those problems that lie within the initial scope. Do not—repeat—do not scale up before you demonstrate success. You will become one of the 85% of pilot failures.
It’s also critical to precisely define the outcomes of your analytics initiative. A wise sage once told me, “You can’t build a road ‘til you know where you’re going.” It was true 20 years ago, and it’s true today. At its core Analytics Ops is outcomes driven. Success is measured by how well the initiative meets your goals. I know this sounds self-evident, but it’s always amazed me that many companies tend to focus on analytics tools, and their capabilities, rather than how those capabilities will actually help them achieve their analytics goals. Tools are not the objective; outcomes are.
Your outcomes should be targeted to the problems you have, they should be quantitatively measurable, and they should inform your decision-making at every point in the process AND shape the processes you embed your analytics capabilities in. Tools should only be facilitators in achieving your desired outcomes. They are a means to an end.
Finally, it’s crucial to involve stakeholders from across your organization to participate in, and champion, the project. Don’t just bring them in for window dressing. Listen to them and value their opinion. They’re on the frontlines, and they know what they need to do their jobs. They’ll play a large part in determining the success or failure of your project.
I’ve only scratched the surface here. Over the coming weeks, I’ll be discussing specific tools and methods you can use to ensure that analytics is embedded within your enterprise—both technically and culturally. We’ll discuss what it means to marry analytics tools with back-end IT and business processes, accelerators you can use to both speed up deployment and ensure success, and a framework you can employ to integrate analytics into your corporate DNA. Stay tuned!