In the course of running IT operations for your business, systems collect a wealth of data data that can yield useful insights to help understand how you deliver services, lower costs and drive more innovation. By analyzing this data effectively, you can get a 360-degree view of the IT business. Thus, the biggest roadblock to getting business insights out of this data is the inability to effectively query for data across these multiple systems.
In the course of running IT operations for your business, systems collect a wealth of data – data that can yield useful insights to help understand how you deliver services, lower costs and drive more innovation. By analyzing this data effectively, you can get a 360-degree view of the IT business.
However, this data is spread across multiple systems and applications ranging from IT service management (ITSM), IT asset management (ITAM), vendor management, project management and financials to name a few. Thus, the biggest roadblock to getting business insights out of this data is the inability to effectively query for data across these multiple systems. Some organizations, for example, use spreadsheets to track different datasets, but this method only allows you to generate the most basic reports, and takes significant manual efforts. Others use the built-in reporting tools within operational applications, but they are limited to a siloed view of data and the reports are rudimentary with basic calculations around sums, averages, means and medians. The result is that IT managers continue to operate with very shallow analytical capabilities.
Defining the data warehouse
Data warehouses serve as the analytical system of record for IT and can help demonstrate the value of IT to the rest of the business. Think of a data warehouse the same way you’d think of a physical warehouse for any major retailer. There are millions of SKUs across hundreds of manufacturers and brands, so the warehouse is very large – yet items are placed and stacked in a highly organized manner that allows warehouse workers to choose items efficiently. Products are organized by categories, such as toys, books, and electronics, and within these categories are further organized by sub-category, price, or even affinity groups. Whenever workers need a particular item, they don’t have to plan in advance. They follow a certain sequence based on the warehouse organization and can retrieve an item easily.
Similarly, a data warehouse organizes information in highly logical ways that mirrors what the business needs to know, and when. Data in a data warehouse comes from multiple systems – such as IT services, finance and call center – and the ability to deal with idiosyncrasies of disparate datasets and correlating them is a key feature of a data warehouse. For example, if the data is received on a daily basis, how will it automatically be organized once it arrives in the warehouse? If records are updated as a result of fresh data, how does this change calculations? What are the different business rules for matching records from different datasets based on common data dimensions such as time, employee, customer and geography?
With a data warehouse in place, your IT organization can gain faster, more flexible views into what’s really happening in the IT department. The benefits are broken down into what we call the “four Cs”:
Capability for ad-hoc analyses: A dynamic and nimble business should not be constrained in the queries it can pose against available data. Sales and marketing teams who use modern Business Intelligence (BI) tools already have these capabilities but most IT teams do not. IT operations managers should be able to quickly get answers to their business questions that cross different systems and datasets, without knowing a priori all the metrics and without waiting for weeks for their analysts to respond. They also need to future-proof their ability to ask the questions they haven’t even thought of yet by being able to run ad-hoc queries against a comprehensive data warehouse. By far, this is the biggest advantage to a data warehouse.
Context: Context is important in our daily lives. For example, when driving on a highway, the posted speed limit signs give us context of whether we are driving too fast. The GPS gives us the context of how far we have come and how much distance remains to be covered. Similarly, an IT data warehouse provides the context about their current situation to an IT organization. For example, a help desk manager would want to know if the week-to-date mean time to repair (MTTR) and backlog of an assignment group is too high or low in comparison to their historical numbers as well as that of their peer groups. What are the maximum and minimum MTTR levels that they have achieved in the past and are they within their average variances? Answers to these questions provide context of whether a team is doing well or needs help.
Correlations: A data warehouse lets you easily correlate data from multiple sources and changes the focus of your analysis to be on business goals, rather than on the topology of operational systems. In order to hold vendors accountable to their SLAs, business owners would need to unite data from a ticketing system with links to assets or the configuration management database, and then again connect it to data from the vendor management application. A data warehouse can correlate data from these sources and enable business owners to identify outages or incidents by each vendor asset, and identify breached SLAs that had business impact.
Calculations: A data warehouse enables calculations beyond the basics of average, count and sum. It allows a range of analytics from percentile, ranking and standard deviations all the way to predictive functions. For example, IT professionals can ask questions such as, “What are the top configuration items that account for 80 percent of our total time spent in incident resolution?”