Embedded BI and Your Data Warehouse
These days, everything we want seems to be available in a disposable, single-purpose version for immediate use right when we need it: paper plates, Kodak FunSaver cameras, one time use credit card numbers for online transactions, even disposable cell phones are easy to find. But what about instant gratification at low cost when it comes to business data management? What if the type of data we collected and the way we used it for business intelligence changed so dramatically that storing it all in a massive data warehouse seemed like a waste of (virtual) space?
By submitting your email address, you agree to receive emails regarding relevant topic offers from TechTarget and its partners. You can withdraw your consent at any time. Contact TechTarget at 275 Grove Street, Newton, MA.
These are questions that have started to arise with the increasing popularity of embedded business intelligence (BI). For a long time, data warehousing has been the answer to BI and related analytics. The more data you have in one place, the simpler it is to mine it for insights. However, this is changing when it comes to some aspects of business decision-making. When you need to make rapid decisions based on real-time data, you don’t turn to your massive data warehouse to pull reports from the last fiscal quarter for review. Instead, you look to embedded BI dashboards and real-time BI reports to provide the insights you need.
How Embedded BI Provides Data
The advantage of embedded BI when it comes to collecting and reporting data is that these functions are housed within the application itself. There’s no need to port the data into another database for number crunching or storage. The information you require can be instantaneously generated from within the application. You can pull spreadsheets, graphic displays of KPIs (key performance indicators), or other reports with a few mouse clicks or keystrokes. This information can play a central role in your daily operational decision-making process.
The information is current – but this doesn’t mean you only have access to recent data. You can incorporate transactional data as far back in the application’s history as you like to review long term trends. As long as the data you want to see is internal to the business application, it’s available to you for use as part of your embedded BI. You have the advantage of knowing that all the information you are looking at is accurate up to the second. You aren’t looking at information in a warehouse environment where data in the parent application may have changed since the last scheduled upload.
It’s Not a Perfect Solution
When you consider the structure of embedded BI, the drawbacks become as obvious as the advantages. If you have many siloed applications that don’t share data in real time, your embedded BI functionality will be very limited. You won’t be able to take a holistic view of your data across your entire organization. That’s why data warehouses still serve an important purpose. They operate as a clearing house and repository for data from disparate sources. In the warehouse, you can compile data from multiple applications and extract the business intelligence you need to make long range decisions. Embedded BI that only provides insight into a single application simply doesn’t offer the scope required for strategic planning.
What data warehouses probably won’t be used for as often in the future is the implementation of “data marts”. That role may become redundant as it can be filled more efficiently and cheaply by embedded BI. After all, why build a data mart that provides outdated information when you can pull what you need directly from the parent application itself?
The Future of Embedded BI
To supplant data warehouses more fully, embedded BI and surrounding/supporting technologies will need to evolve further. These changes will likely be driven by demands from enterprise level business customers. There are a couple of different solutions that may further reduce dependence on traditional data warehousing. First, businesses can upgrade from a legacy mainframe to the kind of interoperable software and data management platform that increases the cross-compatibility of all their enterprise apps. The less data is siloed, the more feasible it is to embed BI technology across or underneath applications. Second, businesses have the option to contract multiple custom embedded BI solutions from a single vendor. An approach that’s tailored to a specific organization’s needs may be able to link resources together “live” so various embedded BI reports can be merged. This type of real-time mashup may become more common as embedded BI matures. In the end, only when data is shared instantly across all applications can organizations use BI for strategic as well as operational purposes.