How many zeros are in a quintillion? Eighteen. (A billion has only nine.) Yes, it’s a dizzying number. But when IBM client Ancestry.com was looking for an enterprise performance management solution for its financial planning and analysis (FP&A) processes, the company needed a tool that could handle its “super-cube,” which held 51 quintillion cells of data.
Ancestry.com is the global leader in family history and consumer genomics. It provides genealogical research and DNA testing that lets people trace their family trees and origins back to specific countries and regions of the world. The company uses sophisticated technology to examine billions of historical records and millions of family trees. The company’s website also offers search tools, family history communities, regional and topical message boards, and more.
Accommodating a super-cube of data
Jose Balitactac, the FP&A Application Manager for Ancestry.com, described the scale of the data challenge the company faces in managing its information. “We have over 10 billion historical records and about 10 million registered DNA participants,” he said. “We needed an application that can spit out the numbers immediately. The old application was taking hours for us to do the calculations and for people to be able to look at the numbers and make decisions.” Check out this video to hear more from Balitactac:
The Ancestry team looked at seven different products before choosing IBM Planning Analytics, powered by IBM TM1. It was “the only one that actually met our requirements,” according to Balitactac. To accommodate the company’s super-cube of data, other solutions would have required them to “break it into smaller cubes, or reduce the number of dimensions, or join members, such as business units and cost centers,” he explained. “We did not want to change our process because our process works for us.”
Calculations done in seconds, not hours
As a test of the solution, Ancestry had the IBM team set up a cube in the U.S. and then Ancestry had a person in Dublin, Ireland, make an entry into the system. They timed how long it took for the model to calculate. “It took less than 10–20 seconds,” recalled Balitactac. That was the kind of response time they were hoping for. “The executives need to have real-time information, and we need to leverage AI and data science. No more of those spreadsheets.”
Now, Balitactac and his team want to expand the use of IBM analytics. They’re looking for ways to bring more platforms and applications into different aspects of Ancestry’s business.
“Do the math” before selecting an FP&A solution
As we mentioned in this blog a few months back, it’s always important to consider a data cube’s ability to hold large volumes of data and for people to “do the math” before selecting a planning solution. To be honest, we were contemplating data quantities several zeros short of a quintillion. But Ancestry’s experience reinforces the value of testing any solution’s data capacity.
Six degrees of separation
If you want to know if you’re an eighth cousin twice removed from some medieval king (or maybe six degrees of separation from Kevin Bacon), Ancestry.com is the place to go. But if you want to learn about planning software that can handle the data volumes and dimensions which make it possible to learn such things, visit our Financial Planning and Analysis page.