Of the companies we’ve studied over the years, we’ve learned that data quality isn’t an outcome or an end-goal; it’s a complex but enriching journey. So far, none would suggest they have reached the end of the data quality process. In fact, many would state they’re beginning or are in the process of an enduring enterprise-wide pilgrimage. At a time when everything is changing, the end state keeps evolving. This being said, in order to become nimble, organizations should mature its data quality practice.
On the other hand, a common scenario that is encountered is when data tells a different story than what the customer holds to be true. Usually, the customer would either blame it on the messiness of the data or the analytics platform itself. After the myriad transformations, matching, schema modifications, unifications and predictive tasks, how can you identify that the data is correct? How can you create a reference data or master data that you can refer to?
We provide recommendations, detailed steps and compositions of the layers needed to design a pre-sales data quality assessment for your customers.
We provide a technical walk-through of the data quality assessment steps and algorithms needed to get a scoring for the quality of the data. This includes data sampling, profiling, scoring, and after evaluation analysis.
We use several industry standards and our very own experience in data quality assessment in order to provide a table of data quality KPIs and formulas. These KPIs can be used as a standard to calculate scores that assess the messiness of datasets.
We build visualizations of a data quality assessment scorecard that presents metric scores to the data stewards observing the business data sets.
We define the relevant Master Data Management architecture.