Fifty-six channels was the actual problem
Lion, a roughly two and a half billion dollar beverage manufacturer operating across Australia, New Zealand, and the United States with about 6,000 employees, had a People and Culture function spread across more than 56 legacy service channels before this overhaul. That number alone explains why routine HR tasks felt complicated to employees long before anyone discussed adding AI to the mix. Lion had already deployed ServiceNow for HR once before this project, but adoption never reached the level the company expected, a detail that matters because it shows a first attempt at consolidation can fail even with the right platform already in place.
The fix came through a partnership with EY and ServiceNow that redesigned employee services across HR, IT, and finance into one experience and workflow layer, rather than attempting to patch the existing fragmented setup. EY ran discovery workshops with HR, IT, and frontline stakeholders and mapped the complete employee lifecycle from recruitment through retirement before any workflow got rebuilt, a sequencing choice that put research ahead of tooling rather than the more common habit of buying software first and discovering requirements during implementation.
Why the first ServiceNow deployment didn't stick
Barak Ravid, EY's Global Alliance Chief Commercial Officer for ServiceNow, pointed to search as a specific, fixable failure point: many employees never even attempted to use the platform's search function because past experience with legacy systems had taught them it simply did not work well. That is a small-sounding detail with large consequences, since a portal employees do not trust to find answers quickly reverts to the fragmented channel-hopping behavior the project was meant to eliminate in the first place. The redesign responded by building multiple ways to find content rather than relying on search alone.
Ravid's broader framing of the architecture is the more durable lesson here: treat the experience layer as the front door and the system of record as the source of truth, rather than conflating the two. Lion kept SAP SuccessFactors as its HR master data system throughout the overhaul and used ServiceNow purely as the unified front end, with DocuSign integrated for e-signatures. That separation let Lion rebuild the experience without touching the underlying data architecture, a lower-risk path than a simultaneous platform and experience migration.
IT support got the same treatment finance and HR did
The redesign was not limited to HR. IT support moved from Cherwell to ServiceNow ITSM as part of the same program, with redesigned service catalogs, consolidated knowledge bases, a standardized taxonomy, improved configuration management database controls, and better request routing all delivered together. Testing happened across offices, retail locations, and manufacturing sites, which matters for a beverage producer where a meaningful share of the workforce works on production lines rather than at a desk, a population IT projects frequently under-serve when design testing happens mostly with office staff.
The project also supported operational and risk management visibility relevant to manufacturing operations specifically, connecting employee service improvements to the kind of operational risk oversight that matters on a factory floor in ways it does not in a typical office environment. That connection between employee experience work and operational risk management is unusual to see called out explicitly, and it suggests Lion's program was scoped with manufacturing risk in mind from the outset rather than treating the factory workforce as an afterthought to a primarily office-oriented redesign.
The numbers that make this worth studying
The results, measured over the first seven months after go-live, are substantial by any standard. More than 95 percent of cases across 49 People and Culture services now run through the ServiceNow portal, replacing the prior 56-plus legacy channels entirely. Self-service usage rose 200 percent, employees accessed IT and HR knowledge resources more than 14,000 times in seven months, and resolution times were cut in half. First-contact resolution rose 87 percent, People Services response speed improved 75 percent, and task success rose 40 percent, with employee satisfaction with People Services reaching 85 percent.
The cost and risk figures are equally notable for a CFO or CISO reading this case study rather than an HR leader. Operating costs across the entire People function fell roughly 30 percent, a meaningful reduction for a function that rarely gets credited with direct cost savings from experience redesign work. An integrated risk environment, built alongside the service redesign, contributed to a 60 percent reduction in active cyber risks, connecting an employee experience project to a security outcome that would not normally appear in the same business case.
Lion's lesson for anyone about to add AI on top of a mess
Early AI pilots using Now Assist and virtual agents reached 25 percent containment within the program, a respectable but modest figure that Ravid's own framing helps explain. His warning, that layering AI on top of an already messy system does not clean up the mess, it just gives the mess a faster interface, lands with particular force given that Lion's first ServiceNow deployment for HR failed to reach expected adoption before this broader redesign fixed the underlying experience. Any enterprise currently racing to bolt AI agents onto an unreformed service layer should read Lion's sequencing as a direct warning.
The six-year transformation program behind these results also moved Lion into a managed services arrangement earlier than originally planned, a detail that suggests the redesign's complexity outran the company's internal capacity to run it alone even with EY's support. For CIOs weighing a similar consolidation, Lion's experience argues for treating the service experience layer as a distinct, fundable project with its own discovery and testing phases, rather than an afterthought bundled into whatever AI or platform initiative currently has executive attention.
