Part III- Four Strategies to Assess Data Quality for Identity Management

March 27, 2015

Identity Management (IDM) systems rely on accurate user data in order to manage user access, enforce policies, and support network security. Historically, organizations have not had a good grasp on the required data profile of their user populations, including employees, contractors, partners, and vendors. In fact, many organizations know very little about their contractor populations today, since they rarely have a single System of Record to manage their contractors; the data simply are not readily available. Data Quality is essential to establishing and managing user identities. Data Quality issues cannot be fixed until they are identified. Here are four simple strategies you can employ to assess data quality:

  1. Define the Minimum Data Set required to establish a User Identity for each user population. As an example, most organizations have some combination of the following attributes for employees:
    • First Name
    • Last Name
    • User type (e.g., Employee, Contractor)
    • Unique Identifier (Note: This field uniquely identifies user accounts across all systems)
    • Colleges
    • UserID
    • Email Address
    • Company/Organization
    • Country
    • Region
    • Manager
  1. Identify and Evaluate your Systems of Record against the Minimum Data Set. A System of Record can be your HR system, Payroll system, Recruiting System, or a partner’s database of users.
    • Employees
      • Human Resources Information System (HRIS) such as SAP, Lawson, Workday
        • Human Resources Information System (HRIS) such as SAP, Lawson, Workday, or Baan
        • Recruiting systems, such as SAP Success Factors or Taleo
        • Other systems (e.g., Office365 for email address)
      • Contractors
        • HRIS Contracting Module
        • Vendor Management System (e.g., Fieldglass)
      • 3rd Party Systems
        • Vendor’s HRIS
        • Partner Database
  1. Map out your User On-boarding and Off-boarding Processes
    • Evaluate data latency & availability - Are the data available when you need them in the process? If not, how long will you have to wait to get the data? Last, can you change the process to improve the data flow?
  2. Evaluate Data across 5 Dimensions (See Part II – The 5 Dimensions of Data Quality)

200 x 200

  • Availability – Are the required data stored in the system?
  • Completeness - How are data collected and how does it move through the on-/off-boarding process and supporting systems?
  • Accuracy – Are the data correct?
  • Latency/Timeliness – Are the data available when and where they are needed?
  • Consistency- Are the data stored in the same format across systems?

More News