If you’ll allow me to make some pretty broad generalizations, I’d suggest that IT people are, in their bones, process-oriented people. Day in and day out, IT people take data from operational systems and move it into enterprise data warehouses and transform that data for operational reporting. It’s their standard “blocking and tackling” and the contents of the data don’t play a major role in their activities.
Conversely, analysts often display significant creative abilities that they apply to their analyses. Their challenge is to take the data from IT and perform sophisticated discovery processes to identify different patterns and trends that were previously hidden in said data. For example, they might sift through data that IT has transformed to identify which customers are good candidates for cross-sell or upsell offers and then build analytical models, using mathematical equations.
Shouldn’t IT and analyst teams fit together well? Shouldn’t their abilities and traits dovetail nicely? Isn’t this an ideal pairing of art and science?
Unfortunately, in too many organizations IT and analysts are an awkward fit. And the reason fundamentally boils down to the time that analysts are sometimes forced to spend outside their core competency: high-value data analysis that uncovers the insights that drive revenue and reduce costs. In many organizations, it’s not uncommon for an analyst to spend as much as 70 percent of his time doing data acquisition, formatting and transformation, and just 30 percent on analytics.
Why is that? It’s simple: They don’t trust the data or the data they really need isn’t available quickly enough. IT, our process-oriented team, is too wrapped up in its repeatable processes to respond to new and disruptive requests. They are focused on the infrastructure and data for a data warehouse, moving data from operational systems to the warehouse every day where it’s used for operational reporting.
The IT person thinks, “If I could just get the data I need out of the operational systems and do the transformations, all at once, with one touch of the data, everything would be perfect.” After all, minimizing the number of touches increases the freshness and relevance of the data. But, that’s largely a vision, not reality.
Instead, the business analysts are accessing information from the warehouse that may or may not suit their needs. So they reach back into the operational data themselves, creating small data marts that create overlap, gaps, redundancies, waste and data problems. There could be issues with a feed somewhere, which the analyst wouldn’t know about. There could be issues with latency or timing that reduce the value of the data, again, unbeknownst to the analyst.
Instead of using his strengths, we have an analyst, who’s been hired to do statistical modeling, spending 70 percent of his time on data acquisition. He’s not able to draw on his training to deliver value in his job. How can we flip those ratios and enable analysts to spend a majority of their time on their core competence?
The answer lies in transforming those operational data warehouses into analytic data warehouses, where the data is better structured for consumption by business analysts across the enterprise, facilitating high-performance analytics. IT can align around a new goal, which is to operate the timely transformation of operational data into locations and formats that are better suited to analytical processing, thus freeing the analyst of these less-suitable IT tasks.
At a strategic level, the answer also lies in communication and alignment. The emerging discipline of high-performance analytics that transforms big data into new waves of high-value insights will not succeed if the organization lacks the proper alignment between IT and the analyst team. IT and analysts share the same goals: intelligent decisions that bring more value to the organization in a stable, reliable, cost-optimized environment. Too often, IT is focused on incremental improvements, delivering a consistently stable, cost-optimized environment. Meanwhile, business leaders are aiming to identify the high-value, high-urgency insights that translate into major advantages. They’re seeking exponential, not incremental, gains. With a little planning and communication, we can help each team understand the other’s goals and achieve a more harmonious partnership.
Gary Spakes is a Systems Engineer Manager for SAS, focusing on both the technical aspects and business implications involving SAS High Performance Analytics.