Loading...

Five steps to getting a single customer view

December 6, 2012 by Guest Contributor

By: Maria Moynihan

The public sector is not unlike the private sector when it comes to data. Both require accuracy and relevancy for optimized processes and decision-making.

For government agencies, maintaining a holistic view of constituents is more important than ever. By linking data across department systems, governments improve operations, citizen profiling and overall record management.  No longer do agencies have to muddle through records of Maria Moynihan, Mari Moynihan, M Moynihan, and other variations of name or contact information when they all are truly one in the same.

Unfortunately, without the right tools and know how, database maintenance, record deduplication, and account validation can be a daunting process.  Below are five critical steps to helping government agencies execute successful linkage of database records:

Step 1: Engage stakeholders

Data stewards are not mind readers. They work with finite data and rely on stakeholders to provide insight. Seek input from users across departments and functions.

Step 2: Identify impacts and priorities

Data errors and disparate data prevent stewards from amalgamating records and defining a master database. Focus on areas of strategic priority.

Step 3: Create success criteria

Look for and set quantifiable metrics for matching. Consider what data needs to be linked and what thresholds are acceptable given objectives.

Step 4: Define new standards

Create established workflows and guidelines for evaluating, merging and purging records.

Step 5: Leverage matching technology

Integrate robust deduplication tools to design multiple workflows and handle a variety of matching challenges.

In short, without data stewards seeking input from commercial stakeholders, an understanding of the data impacts, and establishing a clear process including defined methodologies and technology for deduplication, government agencies will remain challenged in trying to figure out if Maria, Mari, and M are the same person in databases. Click here to see the full guide to Creating a Single View.