Umid Akhmedov is head of data and AI CSA at Microsoft and was formerly the head of architecture and data & analytics at FLSmidth
As a data & analytics leader, I like to create simple goals that anchor data & analytics in tangible business results. While these goals may vary from industry to industry, your topline goals are probably similar to mine:
Of course, as anyone responsible for data & analytics understands, achieving these goals isn’t as simple as it sounds. Huge volumes of data and complex data environments present significant roadblocks.
That volume and complexity can be made even more difficult by the size and history of your organization. For example, FLSmidth is a multinational engineering company based in Denmark with nearly 12,000 employees worldwide. The company has been growing for more than 130 years with numerous acquisitions. Every time a new company is acquired new systems are brought in, new data assets are added that aren’t available to everyone who might need them, and there is a lot of tribal knowledge that gets lost when people leave the company.
To enable data & analytics, first you have to untangle all of that complexity. A data catalog can help immensely by enabling faster data discovery, centralizing documentation on data, and fostering the kind of collaboration that leads to greater productivity. The problem is that while a data catalog solves for some of the biggest pain points in data & analytics, few of these pain points on their own make the business case for buying and implementing a data catalog.
Making the case for a data catalog more often than not rests on your ability to connect the data catalog to tangible business value. In my experience, that means finding concrete use cases that demonstrate how the data catalog can improve the bottom line. I’d like to share a couple of those examples in this blog, and hopefully, give you a better understanding of how to make the business case for a data catalog.
As an engineering company that provides global cement and mineral industries with factories, machinery, services, and expertise, FLSmidth not only makes money with plant sales but with aftermarket parts sales as well. From the time a plant customer makes an order for a spare part, it can take weeks to go from the quoting, ordering, handling, and delivery to getting back up and running. While the customer waits for weeks for their part to arrive, they are losing production time, which means they are losing money. And because of that delay, the customer may turn to local smith, and we miss out on the sale.
Rather than forcing the customer to go through this costly delay or risk losing the sale, we can leverage predictive analytics to identify when a part will fail 90-days beforehand and can proactively inform the customer when they should order the part to experience the least amount of negative impact.
In order to conduct that kind of predictive analytics, we need to understand exactly what data we have and where we can find it — and that’s exactly what a data catalog can help us do.
Another example relates to the kinds of calculations we can do to improve our ability to sell products.
