Phoenix Payroll: A Case Study in Poor Conversion Strategy
The Government of Canada’s beleaguered Phoenix pay system has justifiably gained a lot of exposure in the media over the past year; and we have been monitoring it closely. In case you missed it: in February 2016, the Federal government rolled out a new payroll system called Phoenix intending to save $70 million but instead precipitating a payroll fiasco resulting in underpayment, overpayment, or non-payment to more than 80,000 public servants and student workers nationwide.
All indications suggest multiple points of failure in Phoenix’s implementation—from process errors, data issues and insufficient innovation strategy and IT talent at the government level, to lack of systems training and the sheer volume of government rules regarding wages and salaries.
The fiasco runs deep with ramifications that are still unfolding more than a year later. Last week it was announced that executives in the department heading the payroll integration received just under $5 million in performance pay over the past year. Marie Lemay, deputy minister for Public Services and Procurement, and the public face for the efforts to reverse the errors, has announced a leave of absence for “personal and family reasons”, and nearly 8,000 employees still await payroll adjustments to their accounts. Reports claim it will cost the government an additional $50 million, and counting, in what could have been avoided by a responsible and carefully planned systems transition.
Speaking on the matter, Rosanna Di Paola, the associate assistant deputy minister for Public Services and Procurement Canada said, “We underestimated the time it took people to adapt to the new technology. The learning curve just seemed to be much longer than we expected.”
This was not simply a government project gone awry. From the onset, the integration was undermined by a misalignment of resources and a lack of training that has translated into such issues as students being unable to pay tuition, full-time government employees missing mortgage payments, and the employees who operate Phoenix—improperly trained to run the system—taking stress-leave.
Any sizeable operation needs to approach a change in IT responsibly, not only to protect itself and its employees, but to ensure compliance across the board.
To transition responsibly you have to run the new system at the same time as the old system and test the integrity of the new system. This comparison is a large part of any responsible conversion process and it forms the foundation of User Acceptance Testing (UAT)—a robust analysis of the new system before it is rolled out. Failing to run the systems in parallel and not performing UAT procedures is simply negligent. Training is also an essential part of this process as it is used to ensure your team has the skills and knowledge (and confidence) required to run your new system.
At Metrics, we are knowledge experts in systems transitions and by examining the moving pieces within the Phoenix fiasco, it’s clear to us that what the stakeholders neglected was appropriate testing of the system before it was implemented. They simply couldn’t have performed an audit of the new system with statistical significance before rolling out the system— the errors would have emerged.
This story highlights the cost of rolling out an IT conversion badly. It shows the errors, the tarnished reputations, the employee lives affected, and the money spent afterward in an attempt to clean up the mess. It is clear that spending time and money up-front to create a meaningful innovation strategy and acquire appropriate IT talent to assist with the transition is critical. It also highlights the importance of an innovation strategy—it is much easier to succeed when you have a strategy for innovative success.
At Metrics, we specialize in creating future-focused strategy while mitigating conversion risks. Do you need help transitioning to a new system? We’d love to hear from you: