Image
abstract swish line in gradient colouring from pink to blue on a dark blue background

Integrations - People

Important considerations for people data integrations

There are lots of things to consider when interfacing people data due to its complexity and sensitivity;

 

Why is it needed?

  • What is the data being used for?
  • Is the purpose of the data compatible with the stated business purposes for using employee's data?
  • Has the purpose been assessed by information security?
  • Is the data flow purely outbound or is there an inbound (return) data feed)?
  • What is the purpose of each data attribute? Does it drive process or calculations?
     

What is the population scope?

  • Is it all types of people including contractors, casuals and contingent workers as well as employees?
  • Does it need to include people who are away on leave or otherwise inactive?
  • Which parts of the organisation are in scope?
  • Which locations/regions/countries are in scope?
  • When is a new starter included - as soon as they are known about or only on their start date?
  • When is a leaver included - as soon as the leave date is set or only of the date is past?
  • What changes of existing employees need to be included?
     

What is the data frequency?

  • How often is the data required? (Daily, working day, weekly, hourly, near real time)
  • When in the day is the data required?
  • How long does it take to process the data?
  • What is the timing of the data update and is it automated?
  • Which are the dates that drive the updates (effective date, change date etc)
     

What data is required?

  • Which attributes are required?
  • Are any data sensitive?
  • What transformations are required?
  • How is data to be deleted indicated?
  • How is null data provided?
  • How much data is sent? (per person)
  • What determines which events are included or excluded?
  • How to ensure changes are processed in the correct sequence?
     

How is the data transferred?

  • Is a full set of data provided each time?
  • If only changes are sent - how is the initial data set?
  • How are retrospective transactions handled?
  • How is sensitive data encrypted?
  • How are 'empty files' dealt with - is this an error?
  • What is the method of transfer (API, sFTP etc) and file format (csv, xml)?
  • Is there a need for metadata or reference data to be separately provided?
  • How is the end to end integration monitored?
  • How are errors notified and to whom?
  • How are errors dealt with?
  • What could go wrong and how are risks managed?
     

Common complexities

The following difficult scenarios are quite common where people are concerned;

  • Changes in organisation structures impacting many people at one time
  • Line manager leaves resulting in change to reporting manager in their subordinates
  • Late notification of a leaver - date left is well in the past

Looking for something specific?