The Government of Canada’s HR system currently operates on a classification-based staffing model. The logic of this model is that departments have mandates, and they establish a hierarchy to support outcomes related to this mandate. Jobs are classified and fixed into the hierarchy with generic work descriptions (or in rare cases, specialized work descriptions). Managers fill “boxes” or vacant positions in the organizational chart of their unit, using the classification of the position and its work description to identify the experience required.
This standard GC approach is optimized for a workforce that is predominantly indeterminate. It’s not optimized to serve the recruitment of talent for project-based positions where specific skill sets are required and the intended outcome of the position is time-limited.
At the time Talent Cloud first launched its live platform in 2018, impact statements and the inclusion of key tasks were rarely included in GC job advertisements - both of which appeal to applicants on the hunt for a new position. (Although in 2019 and 2020, we started to notice a few job advertisements on the GC Jobs site including this information.) Our early workshop findings indicated that the inclusion of impact statements and key tasks in a job advertisement would be a benefit in attracting high-performing talent, but how would managers be trained to develop a completely different type of job advertisement? And would HR advisors support the new direction?
Talent Cloud explored the idea that the same logic model supporting departmental structures could be used, in microcosm, to help managers identify the impact individual applicants would have on Canadians. At a grand scale, departments have a mandate, and from this mandate flows intended outcomes. The hierarchy and job classification structure is then organized around delivering these outcomes. To design a recruitment engine optimized for project-based work, Talent Cloud used the same cascade logic model seen in the broader Government of Canada structure, but we shrank it to focus exclusively on the local context of the job being advertised.
Connecting Staffing to Results and Delivery
One of the project’s best “ahah!” moments came from the exercise to refocus hiring around intended impact. In one of the workshops, a manager said she couldn’t actually articulate the impact of the hire she had planned - filling a standard vacancy she had never questioned needed filling. She said she was going to go back to her management team for a discussion about rethinking their upcoming staffing needs, with a new focus on aligning the hiring plan with results and delivery.
To understand how managers and applicants would engage with the proposed model, Talent Cloud held a series of workshops in the summer and fall of 2017 where managers went through a paper exercise for both imagined and real job processes they had planned to test a new model for crafting job advertisements.
Managers were first asked to identify the intended impact on Canadians from the hire they intended to make. We then walked them through an exercise to articulate and prioritize what key tasks would be required to deliver this impact. Once the key tasks were identified, we asked them to articulate what skills would be required for an employee to succeed in delivering each key task. These skills were then summarized and prioritized, and formed the basis of a new way to arrive at the essential and asset criteria for a GC job advertisement.
In the workshops, we tested to see whether managers could complete the task with minimal instructions, and we tested to see if what they produced made sense to HR advisors.
Following the successful workshop testing of the new logic flow, our team created an MS word document which was used by managers to craft job advertisements for live job processes advertised on the platform in the fall and winter of 2018. Refinements were made, and the new process was programmed and became part of the Job Advertisement Builder Tool on the Manager Portal (see the write-up on this in Section 4 of the report.) Managers were interviewed about their impressions of the logic flow and the tools, and reported favourable impressions of the approach.
Talent Cloud also tested the flow of the job advertisement with applicants, interviewing people about their impressions of the usefulness and flow of information provided.
Check out the designs in the Job Advertisement Builder Tool in Research Section 4: Reducing Time to Staff to learn more about how the Impact - Key Tasks - Skills flow looked on the platform.
This component of the Talent Cloud project was an enormous success from a change management perspective. Managers reported that they found the new approach easy to use. This redesign of the job advertisement allowed managers to craft a high quality job advertisement in 1-2 hours. Notably, this was true for both seasoned managers who had learned the standard GC approach for drafting job advertisements, and new managers who had never staffed in government. This showed that not only could the new logic flow be used to change the approach of seasoned managers, it could also be adopted easily by new users.
In addition to being fast and easy for managers to complete, the new process met the approval of HR advisors. Talent Cloud job posters were routinely approved with very little intervention by HR advisors.
Applicants reported that the inclusion of impact statements and key tasks positively influenced their decision to apply when they felt they were a strong fit for the position.
One of the most interesting findings related to this platform intervention was the way in which managers were able to dramatically change their behaviours, easily adapting to a new workflow for producing a job advertisement.
Talent Cloud crafted this new process flow to better optimize job advertisements for project-based work, but there’s no reason that it couldn’t also be used for any job advertisement in the Government of Canada. While Talent Cloud’s platform automates nudges related to optimizing the number of key tasks and skills on a job advertisement, there’s no reason these lessons learned couldn’t be manually adapted by any manager on any platform looking to craft a new type of job advertisement. (That said, to promote a consistent change in manager behaviours and job advertisement content, Talent Cloud found that standardized and automated platform solutions were an important component in delivering results.)
