SOLUTIONS
Human resources functions face a specific automation challenge: the processes that matter most -- onboarding, performance, transitions -- are the ones that most require human judgment, relationship, and context. Automating the wrong steps in these processes does not save time. It creates a transactional experience in moments that should not feel transactional.
HR functions operate in two zones: administrative processes that are highly repetitive and rule-based, and human processes that require judgment, relationship, and sensitivity. The failure pattern in HR automation is misclassifying the second category as the first. When onboarding communication is fully automated, when performance conversations are reduced to a workflow, or when benefits inquiries are routed to a chatbot that cannot handle exceptions, the organization saves time at the cost of the employee experience that HR is supposed to protect.
An HRIS or HR automation platform is purchased and configured before the HR team maps which processes should and should not be automated. The Preserve path is never considered. The tool determines the scope, not the process evaluation.
Onboarding sequences, accommodation processes, and exception handling for benefits eligibility differ by situation and by the HR professional managing the case. Those distinctions cannot be captured in a workflow until they are documented -- and they almost never get documented until an automation project forces the question.
Employee lifecycle events -- hiring, onboarding, role changes, performance, offboarding -- are managed across HRIS, payroll, benefits platforms, and IT provisioning. Automating one without mapping the others creates gaps at transitions that no single system owner ever sees.
HR process work begins with a clear classification exercise: which processes are administrative and rule-based, and which require human judgment and relationship by design? The Process Readiness Score is applied to the administrative category. The Preserve path is applied explicitly to the human category -- not as a default, but as a deliberate design decision. This distinction is what separates a thoughtful HR automation strategy from one that trades employee experience for efficiency.
Before any automation is scoped, every HR process is explicitly classified. Rule-based administrative processes are candidates for automation. Judgment-dependent human processes are candidates for the Preserve path.
The full employee lifecycle is mapped as a connected process -- from offer acceptance through offboarding. Gaps and handoffs between systems and teams are identified before automation is applied to any segment.
Some HR processes should remain human by design. Performance conversations, accommodation discussions, and transition support are documented as preserved processes -- with explicit rationale, not as gaps in scope.
Accommodation requests, leave exceptions, and benefits eligibility edge cases are documented as explicit exception paths -- not left as gaps in an otherwise automated workflow that breaks on anything unusual.
These are illustrative examples based on common patterns in mid-market HR functions. They are not client case studies.
The company had automated new hire onboarding with a sequence of system-triggered tasks and emails. Onboarding completion rates were high. New hire satisfaction scores at 30 days were declining. Exit interviews from employees who left within the first year consistently cited feeling "processed, not welcomed" during onboarding. The automation had optimized task completion while eliminating the human touchpoints that set context and built connection.
Administrative tasks were kept automated. Structured check-in conversations at days 3, 14, and 30 were built into the onboarding as preserved human steps. Satisfaction scores recovered within two cohorts.
Annual benefits enrollment was managed through the HRIS with automated reminders and deadline notifications. Exception requests -- late enrollment, qualifying life events, and dependent documentation disputes -- were routed to a shared HR inbox. Average exception resolution time was 11 days because the inbox had no routing logic, no priority triage, and no escalation path. The automated enrollment was working. The exception process was invisible.
Exception categories were documented, a triage routing layer was built, and resolution time decreased to an average of 3.2 days.
Offboarding was managed via a checklist distributed to HR, IT, and the departing employee's manager. Completion was tracked manually. System access revocation was a separate IT process with no formal trigger from HR. An audit revealed that former employees had retained active system access for an average of 8 days after their last day, in some cases significantly longer.
After the offboarding process was mapped as a single connected workflow across HR and IT, access revocation was automated as a triggered step. The 8-day average was reduced to same-day execution.
The company had implemented a performance management platform with automated review cycle initiation, self-assessment reminders, and manager rating submissions. Completion rates were high. Managers reported spending an average of 8 minutes per review. The HR team received escalating feedback from employees that reviews felt perfunctory. The platform had made it easy to complete reviews. It had made it difficult to have meaningful ones.
The submission workflow was kept automated. A structured conversation guide was introduced as a required step before submission. Manager time per review increased intentionally. Sentiment scores improved.
The following represent common processes in this function that organizations bring to a PFA Diagnostic. This is not an exhaustive list. The Diagnostic begins with your specific situation.
These are detailed walkthroughs using fictional companies. Each follows a real diagnostic pattern, from the initial problem through the DRIFT diagnosis, the Four Paths decision, and the outcome. They are here to show the work, not to replace case studies.
Fictional companies. Real patterns.
Clearview Technology Solutions
Technology services · 350 employees
Clearview's HR team was proud of their onboarding automation. Tasks completed on schedule. System access provisioned within hours. Benefits enrollment triggered automatically. From a workflow perspective, the process was performing exactly as designed. What nobody had measured until the quarterly culture survey was new hire satisfaction at 30 days. The score had declined for three straight quarters. The exit interview data from first-year departures told the same story: people did not feel connected to the organization during onboarding. They felt like they were completing steps. The DRIFT assessment identified the issue immediately: the automation had been scoped against the task list, not against what onboarding was actually supposed to accomplish. The human touchpoints -- the day 3 check-in, the two-week introduction to team norms, the 30-day reflection conversation -- had never been included in scope because they were harder to automate. They were simply left out.
Redesign
Meridian Health Network
Healthcare organization · 1,100 employees
Meridian's annual benefits enrollment ran through their HRIS with automated reminders, deadline tracking, and confirmation emails. For the standard enrollment case, the process performed well. The problem was the exception volume: late enrollment requests, qualifying life events requiring retroactive changes, and dependent documentation disputes that did not fit the standard workflow. All of these were routed to a shared HR inbox. That inbox had no triage logic, no priority rules, and no escalation path. Cases sat based on which HR generalist happened to open email that day. Average exception resolution time was 11 days. Employees calling about their benefits exception were being told to "allow time for processing" by an inbox that had no SLA, no owner, and no visibility. The automated system was reporting 97% enrollment completion. Nobody was reporting on the 11-day resolution time for the cases that required human attention.
Redesign, then Automate
Arbor Financial Group
Financial services · 280 employees
Arbor's offboarding process lived in HR. IT access revocation lived in IT. Both functions had their own checklists, their own timelines, and their own definition of what "offboarding complete" meant. When an employee was separated, HR initiated their checklist. IT received a separate notification, typically from the departing employee's manager, via email. That email was not always sent on the last day. It was not always sent at all. The Operational Truth mapping session uncovered that the gap between HR completing their offboarding checklist and IT revoking system access averaged 8 days. In one documented case, a former employee had retained access for 23 days after their last day. The HR process and the IT process had never been mapped as a single workflow, because they were owned by different departments with no shared system of record for employee status changes.
Automate
Vantara Industrial Group
Manufacturing · 640 employees
Vantara had implemented a performance management platform that automated the entire review cycle: cycle initiation, self-assessment reminders, manager rating submissions, and HR sign-off. Completion rates were the highest in the company's history. The CHRO presented the metric as a success. The culture survey results told a different story. Employees were rating the performance review process as low-value and perfunctory. Manager feedback was that the platform made it easy to submit a review, which had become the definition of completing one. The DRIFT assessment identified the core problem: the platform had been selected and configured before the HR team had defined what a good performance review was supposed to accomplish. The technology had optimized for completion. The organization had never documented what a meaningful conversation required. When the Operational Truth mapping was applied, the HR team identified that the platform was designed as if the submission was the output. The conversation -- the part that required manager judgment, specific feedback, and genuine engagement -- had been treated as something that happened before the platform, not as a step that needed to be designed into the process.
Redesign, then Preserve
The Preserve path exists for a reason. The DRIFT Self-Assessment helps identify where your HR automation environment has gaps -- and where it may have gone too far.