Z-Codes, a technical name for bad data

1 October 2, 2014 at 9:58 am by

Since 1997, CSIO (Centre for Study of Insurance Operations) has been transiting data from insurers to insurance brokers via CSIOnet. This daily download process is now part of daily life for brokers, receiving all-important policy data from insurers and transferring that data into their broker management system (BMS) client accounts. The process is deceptively simple. Insurers process requests for new business, endorsements and renewals from brokers and then send that data back to them using CSIO-approved codes in a batch process via CSIOnet.


CSIO has been the ‘keeper’ of these codes for the past 30 years, using a process from ACORD (United States standards association) that has since been licensed to CSIO.

The process of creating codes has evolved over the years, but recently has undergone a sea change. The original set of forms and codes took months to negotiate between all the stakeholders, namely the insurers, brokers and BMS providers. Since the beginning of the data transfer process, there has been a form and code change process in place that is inefficient and difficult to manage, taking upwards of eight months to process simple additions to the code book. This is in addition to the process of revising and adding new data fields to forms and applications, as evolving needs have dictated.

This process inefficiency led to the creation of a parallel coding system, known as Z-Codes, to allow insurers to rapidly insert changes into the code book without going through the code approval process. Insurers became accustomed to using Z-Codes as the preferred methodology for releasing new endorsement codes, rather than submitting endorsements for approved codes and having a strategy for the migration to proper CSIO codes. The result of this has been a significant deterioration in the quality of data that is sent from insurers to brokers via CSIOnet.

Brokers can verify the presence of Z-Codes in their data sets by looking at the endorsement fields in their habitational and automobile insurance applications in any client file. If their experience is like mine, they will find that upwards of 70% of the codes are Z-Codes, even though appropriate codes exist in the CSIO library for most, if not all, of these endorsements. As individual insurers create different Z-Codes for similar endorsements, brokers cannot easily search their systems to find clients with similar endorsements across insurers. An example of the need for this type of activity was the recent issuance of a new policy form in Québec for Replacement Cost Automobile Insurance the QPF 5. Brokers wanting to search their databases to identify all their clients who have this coverage on their existing QPF 1 policy would have to have the Z-Code equivalent for every insurer’s QEF 43A&E so that they can know which clients they should contact to offer the new QPF 5 policy.

14 months ago, under the direction of Hans Gantzkow, Senior Architect, CSIO changed the code approval process to be much more streamlined and user-friendly. Where requests for new codes used to take up to eight months for issuance, code requests can now be processed in as little as three weeks. This improved process should motivate insurers to change their implementation processes to include the creation of approved CSIO codes at the outset of their product development, as opposed to making requests post-hoc. The result will be cleaner, better quality data that brokers can use to improve their customer service and market segmentation processes.

There is an immediate need for brokers to request insurers to undertake a process to review, update and renew their codes. This will ensure the quality of data that brokers have in their BMS conforms to CSIO standards and allows them to make proper use of the data for their own business analysis.

Note: By submitting your comments you acknowledge that insBlogs has the right to reproduce, broadcast and publicize those comments or any part thereof in any manner whatsoever. Please note that due to the volume of e-mails we receive, not all comments will be published and those that are published will not be edited. However, all will be carefully read, considered and appreciated.