Onboarding redesign

The first 90 days across 37 locations

A new hire could complete orientation, begin role training, and still depend on a manager to explain what came next or where to find an answer. The redesign connected orientation, role-based learning, manager sign-offs, the internal AI knowledge assistant, and 30-, 60-, and 90-day follow-up.

412new hires evaluated
38%faster certification
+17 ptsfirst-attempt pass rate
+11 pts90-day retention
Evaluation designSix months before launch compared with six months after launch · 412 new Guest Experience Guides · 37 corporate locations

Before the redesign

Training existed, but the learner still had to navigate the gaps between it.

Orientation, role learning, manager coaching, and operational answers were handled separately. Completion did not always happen on time, and readiness depended heavily on the local manager.

New employees could remember that a topic had been covered without knowing where to find the answer when the question appeared during live work.

My role

My role covered the full 90-day pathway.

Needs analysisreadiness gaps, learner needs, manager workload, and completion failure points
Program designorientation, role-based learning plans, certification standards, and follow-up
Deliverylearning assets, manager sign-offs, LMS assignments, and AI knowledge support
Evaluationcompletion reporting, certification results, Day-30 readiness, and 90-day retention

Learner journey

The first 90 days

01

Day one: orientation

Revised orientation introduced the guest journey and the operating expectations that applied across roles.

02

Before certification: role learning

Each learner received the modules, practice, and standards tied to the work they would perform.

03

Certification: manager sign-offs

Managers recorded observed skills against defined criteria instead of relying on a general confidence judgment.

04

First 30 days: answers during live work

Apollo, the internal AI knowledge assistant, let new hires retrieve approved guidance without knowing how the source library was organized.

05

Days 30, 60, and 90: reinforcement

Scheduled practice, validation, and retained-knowledge checks continued after orientation.

Manager workflow

What changed for managers

Managers received the role plan, the required LMS assignments, and the criteria for each observed sign-off. Completion was recorded against specific skills rather than a general judgment that the learner seemed ready.

The 30-, 60-, and 90-day checks also gave managers scheduled points to revisit performance after the initial orientation period.

Before-and-after evaluation

Results across 412 new hires

Time to role certification
24 days→15 days
38% faster
First-attempt certification
76%→93%
+17 points
Manager-rated Day-30 readiness
6.4→8.7
+36%
Questions requiring a manager
14.2→6.1
57% fewer
Process-confusion incidents
3.8→1.6
58% fewer
90-day retention
76%→87%
+11 points

First 30 days

Finding an answer during live work

Eighty-eight percent of new hires used the internal AI knowledge assistant during their first 30 days, and 74% of onboarding questions were answered without manager intervention.

4m 40s → 22smedian answer-retrieval time
1,154 hoursestimated manager time returned across 412 hires
92%positive answer rating

“Apollo has become the part of onboarding that continues after onboarding ends… It has made training feel less like a one-time event and more like a resource that stays available throughout the employee’s development.”

Store Manager

Retention pattern

Completion and 90-day retention

Full pathway completed on time91%90-day retention
No structured completion66%90-day retention

The full pathway included orientation, the role-based learning plan, and at least five interactions with the internal AI knowledge assistant during the first 30 days. The difference is an association in operational data, not a randomized causal experiment.

Evaluation limits

What the data does and does not show

  • Results compare the six months before launch with the six months after launch across 412 hires at 37 corporate locations.
  • The intervention changed several connected elements at once, so the outcomes are attributed to the onboarding system rather than to one course or tool alone.
  • Manager-time figures are estimates based on observed question volume and retrieval time.
  • A separate annualized business-impact model exists, but modeled dollar value is intentionally excluded from the headline claims on this page.