Embedded analytics – Dig a little deeper with embedded BI
Embedded BI (business intelligence) is making a splash as the latest and greatest way to leverage data for operational decision making. But there’s one thing that the proponents of this approach to real-time data reporting don’t necessarily mention. The dirty little secret is this: Your embedded business intelligence is only as smart as you are.
That’s right. Embedded BI will not help you connect with customers, speed products to market, increase profitability, mitigate risk, overcome obstacles, promote better choices or point the way to your company’s future – unless you are doing it right. That’s why it is critical to move beyond the idea of embedded BI as a set of reports or tools and into the deeper principle of embedded analytics. After all, without critical analysis, information is just noise.
How does embedded analytics differ from embedded BI?
At first glance, these might seem like two different terms that mean basically the same thing. However, the practice of embedding analytics into both business processes and business applications is actually the more foundational concept. Embedded BI is the ideal end product – a set of actionable insights gleaned from the real time reporting and appropriate analysis of business data.
Embedding analytics means asking the right questions
In order to incorporate analytics into business processes and enterprise software, an organization must look beyond the way they currently aggregate and interpret data. Yes, the information you are already collecting with embedded BI tools is important and a good starting place for experimenting with embedded analytics. However, in order to truly revolutionize how your business operates, you need to start asking even more questions. These queries fall into several categories. Of course, we’re assuming you’ve already determined the short term operational objectives and long term strategic goals for your business as a starting point. Those goals serve as the framework within which to answer all the following questions:
Data quality, availability, and use
- What data are we currently collecting via embedded BI and other tools? Is this the data that we need? Who decides what we need to know?
- Is there a set of key performance indicators we could be collecting and analyzing that we are neglecting? How might we figure out what data we are missing (blind spots)? Can we drill down into each set of data far enough to glean all important information?
- What are we currently using this data for? Are these uses in alignment with our goals? Can we make comparisons in performance/outcome (one of the most important types of analysis) based on the real-time data we are collecting?
- Is the data we are collecting correct and complete? What checks and balances are in place to ensure this?
- Are we collecting data at the right point in the workflow? Would collecting it earlier provide more time for decision making? Would collecting it later provide a more accurate picture upon which to base a decision?
- Do we have control over each source of data, or are we dependent on third parties for any information? If so, what impact does this have on timeliness and accuracy of this data?
Existing embedded BI tools
- How do our various BI tools interact? Is there an overarching program that permits us to use these tools across applications or is each one simply a standalone product?
- Do our employees understand the purpose and importance of our embedded BI tools? Do they know how to use them correctly to support business objectives? Are they actually using the BI tools as intended? How are we tracking this (that’s where embedded BI goes “meta”).
- Are our embedded BI tools easy and intuitive to use with visual representations the average user can understand? If not, how can we improve them?
- Is there a process in place for re-evaluating and refining our embedded BI tools as circumstances change?
Business process evaluation
- Are there any core business processes that don’t lend themselves to scrutiny with embedded BI and analytics? If not, why? What changes are needed to enable embedded analytics in these business processes? What would the ROI be if such changes were made?
- Are any new core (or peripheral) processes being designed for our business? How can analytics capabilities be structured into these processes from day one?
- How will making decisions based on embedded analytics affect existing processes and workflow? What challenges might this create?
- What aspects of embedded analysis should be automated when it comes to directing the flow of business processes and tasks? Which aspects should be done under human supervision (or trigger an alert) for real-world or common sense decision making that a computer can’t handle?
As you can see, these questions are important to answer whether you intend to overhaul your existing processes yourself or hire a consulting firm to revamp everything for you. It’s a lot to think about, but the results are worth it. By the time you’re done analyzing how your organization currently operates with embedded BI and where you want to go with embedded analytics, you’ll have a deeper understanding of your business than ever.