Whether we are at conferences or giving online product demonstrations, one of the questions we get asked the most is: “Can I do impact analysis with MapWorks?”
The answer is a resounding “Yes!” and for good reason; impact analyses is one of the most important use cases for metadata management platforms. Being able to take a data element and very quickly see: (a) what upstream sources affect it; and (b) what downstream systems if affects is where metadata tools can really shine. Data teams spend a lot of time entering data elements and making sure each is mapped appropriately. For large systems with thousands of data elements, teams will have invested thousands of hours documenting definitions, formats, and mappings. The payback on that investment comes with a few clicks of the mouse, when everyone can see how changing something affects everything else.
Here’s an example that we use in our customer demos.
Consider a bank that uses four different systems for managing its loan accounts. It has a system for commercial loans, one for consumer/retail loans, one for construction loans, and one for credit card accounts. Each of these systems has certain features for servicing these different loan products, but the systems have things in common as well. Every one of the bank’s loans has an account number, an open date, and interest rate, and a balance (amount owed by the customer to the bank). Additionally, loans on each of these systems have an “Unfunded Commitment Amount” that is, an amount that’s left for a customer to utilize at some later time. Think of a credit card with a $1,000 balance and a $5,000 credit limit. The $4,000 difference between the limit and the balance is the “Unfunded Commitment Amount” that the customer can utilize for purchases.
Loan balances and unfunded commitment amounts are the key metrics used in most of the bank’s reporting, but executive management and bank regulators don’t really care which system is used to service a loan. They need to aggregate balances and unfunded commitments across these systems by line-of-business, product, geographic area, borrower’s industry, collateral type, customer type, customer relationship, or lending officer. In our example, the bank has a data warehouse that’s used to integrate data from all of these systems into a common set of data that can be used for reporting and analysis. From the data warehouse, integrated data is sent to downstream data marts and ultimately reports, dashboards, predictive models, etc.
Once these data elements and relationships have been mapped in MapWorks, you can visually see where the data comes from and where it goes. In our data menu, go to the UNFUNDED_AMT element in the data warehouse folder. Then choose the LINEAGE tab to see everything upstream and downstream from that element
This is a simple example of the data lineage you would expect. In large banks, it’s not uncommon to have 40 or 50 individual source systems mapped to a single table/field in a data warehouse or reporting tool. In both simple and complex cases, the MapWorks lineage tool shows the the impact of upstream or downstream changes.
Finally, we have a System Report in MapWorks that shows field-level detail for every system upstream and downstream from the system you want to view. If you want to see every field from the Commercial Lending system that affects the Data Warehouse, just run the System Impacts report for the Data Warehouse and you’ll get a detailed analysis of everything going in and out of it.
The Systems Report can be printed or you can copy/paste from it for further analysis.
Your metadata platform is more than just a place to store data definitions and formats. The right tool with the right metadata gives you a powerful way to understand the impact of changes to your data environment.