How Yellowfin controls its data quality

How Yellowfin controls its data quality

When building and running an organization, it's critical that you trust your data. A statistic that shocked me recently was that 84% of CEOs don't trust the data they’re basing their decisions on. While it’s a common problem, it’s one that’s actually easy to fix.

It may sound obvious, but in order to trust your data you need to control your data quality. That is a simpler process than you might think. At Yellowfin we follow four key steps to ensure the integrity of our data and any organization can adopt them.

1. Understand your data sources

Data quality control starts by understanding all your data sources. Where do you get your data from? What data is stored in there? Make sure you and your team understand the rules of that data and how it works.

2. Centralize your data

Once you know your data sources, you can extract and centralize them. At Yellowfin, we put our data into a centralized data warehouse environment. The team in charge of this process works closely with the operational data owners. Take finance for example. There's a constant dialogue between our finance team and our data warehouse team about the numbers we're taking out of the finance system.

3. Reconcile constantly

We reconcile our data every month. That means that we cross-check that the numbers in our data warehouse are exactly the same as the numbers in our operational systems. Again, to use finance as an example - the finance team work with the data warehouse team to make sure that all the numbers are aligned, accurate and able to be explained.

4. Maintain one source of truth

Once you’ve done your reconciliation from a centralized environment and have sources you understand, this becomes your source of truth. Every piece of analysis we do at Yellowfin comes from our data warehouse. That means every number that we create, use internally in presentations or Stories, or send externally comes from the data warehouse. We know we can trust our data there.

That’s our entire process to ensure our data quality. You can extrapolate this process out in large organizations, but it works equally well as a philosophy and process for a small organization.

There's no magic to it, it’s just a matter of dedication to the process. All the tools you need to implement this process exist. It's just a matter of making sure you put the processes in place to validate your data and continue to emphasize data quality as part of your culture.  Once you know you can trust your numbers you can analyze and understand how your organization works and make decisions around that data. This can make a dramatic difference to your business.