Practical lessons from a real multi-country HR transformation: process harmonization, behavior change, and where AI actually helps.
Key Takeaways
- Why process/policy harmonization must precede BPO—and how to phase rollouts across 20+ countries
- Change fails without behavior shifts: train the “new way of working,” not just the steps
- Rehearsals in 3D: breakouts, regional squads, live resistance mapping
- Scenario room win: leave-of-absence flow + RACI + knowledge check
- Measure adoption weekly; target pockets still routing work to HR
- Near-term AI value: leader practice reps, scaled employee voice, faster feedback loops
Outline & timestamps
- 00:00 Intros and Introduction to Change Management
- 04:00 23-country HR BPO phasing
- 08:00 Behavior change & ways of working
- 13:00 3D rehearsals
- 18:00 Scenario room (US LOA)
- 21:00 Adoption metrics (ServiceNow)
- 24:00 AI for leaders (practice + feedback)
- 29:00 AI for employees (anonymous voice)
- 36:00 HR × IT ownership, budgets
- 39:00 Next 3–5 years
- 46:00 Wrap
Speakers
Yvette Oliger — Senior Director, Global Benefits & Compliance; led a phased HR transformation across 23 countries.
Angie Kester — Founder & President, Attune Collective; org design & change lead (rehearsals, micro-sessions).
Danny — CEO, Hyperspace; host and facilitator.
Panel Session Recap
On this panel we walk through Glanbia’s 23-country HR transformation: harmonize processes/policies first, phase go-lives, then train WoW (ways of working). We show how large-group rehearsals in Hyperspace surfaced resistance and shaped training. A scenario room for US leave of absence clarified roles (manager/HR/carrier) and ended with a quick knowledge check. Adoption was tracked weekly in ServiceNow to spot pockets still routing to HR. Finally, we cover where AI helps now, leader practice reps, scaled employee voice, and why ‘pruning’ beats blanket AI rollouts.
*Note: Content pulled from live discussion, only lightly edited for clarity.
Intros & what the session covers
What this is: Real change management, not theor; where process meets behavior.
Who’s talking: Yvette Oliger (Glanbia, Global Benefits & Compliance; led HR BPO), Angie Kester (Attune Collective, org design & change), Danny Stefanic (CEO, Hyperspace; ~30 yrs HR)
Yvette’s HR Transformation Journey (Glanbia)
- Goal: Outsource to BPO, but only after harmonizing processes & policies
- Scope: 23 countries, phased (14 + 9). ~130 processes documented; HRIS configured & tested
- Local variance: Standard global flow with country-specific steps (e.g., works council in Germany)
- Behavior change: Train Ways of Working (WoW) so employees use ServiceNow/self-service instead of routing to HR
- Reality check: Without change management, harmonization work is wasted
- Status: Phase 1 live (June 30); more training + continuous improvement through 2026
Transcript TL;DR (30 sec) Yvette Williger on Glanbia’s Transformation Process
Glanbia is outsourcing to BPO after harmonizing processes and policies, then aligning HRIS and training Ways of Working so employees use ServiceNow/self-service instead of routing to HR. The rollout spans 23 countries in phased waves (Phase 1 live June 30), with ~130 processes documented and country-specific steps where required (e.g., works council). Enablement continues post–go-live, with ongoing training and continuous improvement through 2026.
We can do all this work. You can do the harmonization, the standardization, 130 processes, what have you, but if you don’t manage the change, all that work will be for not. – Yvette Williger
Full Section Transcript: Yvette Williger – Glanbia’s Transformation Process
The goal was to outsource and move to A BPO (Business Process Outsourcing) in order to do that. What is highly recommended is that you ensure you have your processes harmonized and your policies as well. Danny mentioned 23 countries. That is correct. We and we didn’t do it all in one full swoop. So this is a project that was done in phases.You had a phase one where we did target. 14 of the countries, and then the remaining 13 or the remaining difference will be going live by the end of the year. What we did is the first thing we did is to make sure that we had all our processes documented, which we did not. But we did finally get to that point.
Quite honestly, we have about 130 processes.
The shared services model that we had, service UK, Ireland and the US which is predominantly where our folks are located. Knowing that those processes were harmonized and standardized for the three major countries, we then proceeded to meet with the remaining.
Countries, China, Sweden UAE, Netherlands France, India.
So what we did is we then met with each of the country selected representatives. Within human resources to review the processes and to ensure there were any, if there were any call outs. If there were, then we then knew that we would have to create a sub process.
I’ll give an example. Hiring an employee. So it would be maybe we have a standard process that maybe in Germany it’s done slightly different, which actually we do have an additional step due to the works council. So we made sure that we documented that as well. While I had a team working on that, I also had a separate team working on the policies.
We collected all of the policies. We focused on LEAP. The LEAP policies are quite extensive, and so we partnered again with an HR representative. Once my team completed the review of the policy to ensure that it was compliant, then we met we sent that over to human resources. For one final look and sign off, which then transitions us to the next step.
HR HRIS, the configuration. I had a team that worked on the configuration and also on the testing before we went live. The change piece is something that. That can really make it or break it. And I’ve said this before to Danny, I’ve also said, refer to Angie and to many who have actually worked with me: We can do all this work. You can do the harmonization, the standardization, 130 processes, what have you, but if you don’t manage the change, all that work will be for nothing. We did partner with Angie [Keister]. She was on the team and we then mapped out how do we communicate the change. And there are two ways to get this done.
So you have to do both. One is to ensure that you go over the training. So we sat with them, we reviewed the process. We called out any step outs, which to be honest, there weren’t that many, which is great. But once you do that, you also have to then circle back and conduct a training. So what is that new way of working? And we did a lot of Ways of working (WoW). And then the behaviors, which is really where this tool [Hyperspace Metaverse Platform] really came into play. What are the behaviors? Now an example we also, by the way, implemented ServiceNow, which is where then the policies are stored, but the behavior of now you have an employee who will use this tool as opposed to going to human resources when they have a question.
There were many changes, especially to what we call the subscale country. Non US, UK, Ireland because they were used to doing the paperwork onboarding manually. A new hire, which is now being done completely differently. Automation. So we are not done with the project. It’s we actually went live with phase one on June 30th.
And as I mentioned previously we will be going live with the remainder of the countries by the end of the year. More training. I will say that the change is ongoing. Change Management. It’s not something that’s gonna stop. It has to continue. And evolve. And then the last step is then the continuous improvement, which is also something near and dear to my heart. But as we continue not to work with these processes and we continue to evolve, it will give us all the opportunity to call out any improvements. That’s also something else that we will be focusing on for 2026.
Angie Kester — Large-group rehearsals (employee engagement → targeted training)
Angie ran seven large-group sessions in Hyperspace (tables, prompts, breakouts) to surface resistance and unknowns fast, then used that input to retool training. She aligned leaders first (executive + director micro-sessions), put standardized processes in-room so regions could discuss real changes, and closed each wave with a quick knowledge check. Net: employee-led insights, faster adoption, better WoW enablement.
“Change lands when it’s leader-led but employee-engaged.”
Full Section Transcript:Angie Kester – Rehearsals & Enablement
Project context
I was on the project for a year helping with change management. There were many moving parts, as Yvette described. A global team was managing the technology and other departments; our focus was HR, but several functions were going through transformation. As part of the larger change team, my responsibility was to help HR make the jump—align the HR leadership team, define the org-design changes needed for the new operating model, and ensure leaders knew how to communicate and lead their teams.
Leader-led, employee-engaged
Change is most successful when it’s leader-led and employee-engaged. HR employees were involved in the process harmonization and standardization Yvette’s team led, along with other stakeholders and representatives. Change is also about learning and unlearning—our ways of working.
Large-group sessions in Hyperspace
We constructed large-group sessions using the Hyperspace platform. We set up tables as we would live (five people per table), with a screen at the front and another at the back of the room. People could interact and—even at different computers—felt a sense of being together. With Hyperspace’s help, we set up guided discussions and breakouts. I provided questions (for example, changes to processes, what people wanted to know more about, changes they noticed). We added the harmonized and standardized processes to the room so regional teams could interact with them.
What we learned / why it mattered
The engagement, dialogue, and questions let us hear where resistance was and understand the unknowns. That input helped us tailor the intensive training that followed.
Scale and cadence
We ran seven sessions over two weeks, with about 25 people invited to each session. Turnout was good, and it was a big effort to involve everyone. Traditional learning-and-development training followed the sessions.
Closeout moment
At the end of sessions, we met on the rooftop, threw emojis, and took photos—then returned to day-to-day work.
Leader micro-sessions before rehearsals
Before the rehearsals, we held three micro-sessions with the HR executive team and mid-level leaders (directors and above) to answer questions on tricky topics we knew could become holdups. These followed executive alignment sessions. After rehearsals, we provided more specific employee training as waves progressed.
Why additional training
We saw the need not only around the case-management technology for the centralized service team, but also in some process areas so HR business partners clearly understood what was changing. Some had already been following the standardized, centralized way; others had not. That added complexity required targeted training.
From Table Talk to Adoption: Yvette’s LOA Demo & Weekly Dashboards
They ran a full LOA scenario in Hyperspace with avatars per swimlane to nail role clarity (manager vs HR vs carrier), then did a quick knowledge check to confirm understanding. Adoption was tracked via ServiceNow dashboards weekly to spot pockets still routing to HR and course-correct. The group entered the room together but progressed self-paced; Phase 1 went live June 30 and change enablement continues into continuous improvement.
From Table Talk to Adoption: Yvette’s LOA Demo & Weekly Dashboards
Harmonized LOA was acted end-to-end with avatars per swimlane, locking role clarity (manager / HR / carrier), followed by a quick knowledge check and weekly dashboards to track real adoption.
[Scenario room walkthrough]
One of the processes that had been a real headache in the US was the leave of absence. It was being done very differently throughout the US. We were able to harmonize that process and have a standard policy. Using the platform, we had an avatar for every swim lane owner and ran through the whole case end-to-end.
[Role clarity outcome]
The play-through made clear when the manager steps in, when HR steps in, and when the employee should reach out to the carrier directly. We had an avatar for every single swim lane owner, with the process visible in-room and on screen.
[Knowledge check]
After the room, we regrouped and ran a quick knowledge check to see how much participants had gained. People voted by emoji—nothing scientific, but it confirmed understanding before moving on.
[Self-paced vs group flow (clarification)]
The whole group entered the room at the same time, but each person progressed at their own learner pace once inside, interacting with the avatars and process individually.
[Measuring adoption (ServiceNow)]
We implemented ServiceNow and ran dashboards/metrics weekly. The reports highlighted behaviors—whether employees were going directly to the right channel or if items were still being routed to HR. Where pockets weren’t following the model, we set up calls to address and change the behavior. Month-by-month tracking showed change as communication and training landed.
[Early challenges & trajectory]
The beginning was not easy; this is a significant shift, especially for regions moving from manual onboarding. Phase 1 went live June 30, and we began to see good numbers afterward. Change management continues alongside rollout and into continuous improvement.
Danny Stefanic – AI’s role: practice, feedback, and scaled listening
Use AI to practice tough change conversations safely, get instant feedback, and vary objections until you’re ready. Pair that with anonymous AI chats to capture real employee sentiment between surveys. Because AI keeps accelerating, the real work is change management – building comfort with continuous change and focusing on outcomes over effort.
Danny Stefanic on AI’s role: practice, feedback, and scaled listening
Leaders get safe reps with realistic objections and instant feedback; employees get an always-on way to be heard. Surveys still matter—but AI conversations surface richer signal and reduce fatigue.
[Leader practice with AI]
As a manager, introducing change can feel risky. With people-centric AI you can talk to “team-like” representatives, practice how to communicate the change, hear objections, and see how you’d handle them—before meeting the team. Each discussion can be different, so you prepare across a range of challenges.
[Immediate feedback]
The value isn’t just the natural conversation; it’s the near-instant feedback from AI characters on what you did well, what missed, and what to try next round.
[Employee voice beyond surveys]
Surveys are effective, but there’s survey blindness—you can’t send one every day. Short AI conversations let people share how the change is landing, in their own language, at their own moment. AI can review and consolidate the findings—even anonymously—so you can spot sentiment and ideas without putting people on the spot. Think of it as a “corporate therapist” that listens at scale.
[Accelerating change = change management challenge]
AI capability is advancing week-to-week. That pace is itself a change-management problem: organizations need people comfortable with continuous change, and processes that measure value by outcomes, not effort.
[Ownership and next steps]
Adoption sits across HR, IT, and leadership: get tools in hands (e.g., Copilot), keep listening, and iterate. We’ll continue exploring where AI fits best in learning, training, and change work.
Ownership & Enablement — HR × IT, Budget, and Timing
HR and IT co-own adoption. Stabilize first, then fund what has a business case. Early tools (e.g., Copilot) are in play; broader AI rolls in step with enablement and continuous improvement into 2026.
Ownership & Enablement — HR × IT, Budget, and Timing
AI/change adoption sits across HR and IT: stabilize, enable, then iterate. Budget follows a clear business case; 2026 planning is about continuous improvement, not a one-and-done rollout.
[Who owns what]
HR partners with IT to evaluate, prioritize, and fund initiatives; leadership alignment determines pace and scope.
[Near-term reality]
Tools like Copilot are in early use; broader AI use is being explored with IT to ensure the case is there before requesting budget.
[Roadmap]
Stabilize current waves, continue enablement, and push continuous improvement through 2026.
Mindset: prune, don’t blanket-rollout
Be surgical: deploy AI where it clearly helps, kill the rest. Train, measure, and review to protect quality and outcomes.
Mindset for AI in Change: Prune, Don’t Blanket-Rollout
Treat AI as selective enablement, not a monolith. Double down where it amplifies strengths; quickly cut what doesn’t add value.
[Why pruning matters]
Blanket rollouts spread resources thin and risk quality. Different functions (finance, HR, IT) need different AI fits and guardrails.
[Quality over breadth]
Precision, training, and periodic checks prevent “hallucination-driven ops.” Focus on outcomes, not tool count or coverage.
Danny Stefanic: Continuous change & outcomes-over-effort
AI is forcing change management to grow up: the pace won’t slow, so leaders need safe spaces to practice hard conversations, get instant feedback, and iterate. Employees need ways to be heard between survey cycles. Used well, AI complements surveys, surfaces quiet champions, and keeps focus on outcomes over effort, making adoption continuous and behavior-led. Practice, listen, prune, while tools evolve week to week.
Danny Stefanic on Continuous Change: Competing on Outcomes, Not Effort
AI’s capability is accelerating week to week. That pace is a change-management challenge in its own right: we need people comfortable with continuous change and orgs that measure value by outcomes, not hours.
Why this matters
The old pattern—one upheaval, then a new steady state—doesn’t hold. Tooling keeps shifting, so adoption must be iterative and behavior-led, not a one-and-done rollout.
What good looks like
Frequent, small experiments; fast feedback; pruning what doesn’t add value; enabling teams to improve their own ways of working.
One of the challenges I see with AI is that it’s accelerating so quickly, and that in itself is a change-management challenge, where the technology is becoming able to do new things month after month, sometimes week after week.-Danny Stefanic
Danny Stefanic- Scaled Listening: Surface Champions & Ideas
I think the key to all this really, if I’m understanding correctly, is listening to your workforce—being able to listen. We can do that as great managers or champions in the organization who have the ability to listen really well and try and bring that into the organization’s awareness. But using AI in the way you’re describing allows you to do that at scale.
And it may even allow you to highlight and get visibility on new ideas being contributed by people that you didn’t expect, or uncovering champions of change across the organization who are just itching to contribute but may not have the loudest voice. Through these AI discussions and analysis, that might bring these opportunities more to the forefront.
Short, anonymous AI chats capture real language and context between big moments. Patterns reveal friction, and unexpected champions you can amplify. Use the signal to target training, comms, and quick wins.
Danny Stefanic- Surveys and AI Chats: Complementary Signals
Surveys are really effective. But there’s a certain point where there’s survey blindness—you can’t send one every day. You have to time them out so people don’t just fill them out without consideration.
I think that’s where the AI conversations can really shine: you get a chance to really engage someone; they’re paying attention; they can be relatively short; and through that conversation more value can be extracted than you may get in a survey.
What we’ve been doing with learning and training, we can say what we want the scenarios to be, and the AI will generate a scenario that does that challenge based on the rubric or outcomes we’re looking for—communication skills, empathy, anything like that—so the same can go with surveys. It’s not as finite as surveys, but you can extract more value, and that’s the benefit of AI. And AI can review and consolidate the findings—even anonymously—so you’re not necessarily holding anyone accountable for how they feel and making them feel threatened about that. So anonymous AI can listen—basically your corporate therapist—everything stays confident, but you’re able to express how you feel. What we used to have is an upheaval—some technology or a new way of doing business or even policies and regulations—then we’d adapt and find our new comfort zone. Looking at the rate of change in AI, I believe we’re in for consistent, accelerating change.
Organizations may need to stay ahead to have competitive advantage. You may not be able to charge for your time anymore, but really for your outcomes—how quickly you can use these tools to create real value. If things are going to change consistently, then change management becomes an even more critical part of nearly every business.
I see the role of change management and AI accelerating—people need to be taught to be comfortable with consistent change. That’s creating a new type of mindset, and I think that may be the challenge we’re looking at in the next three to five years for people skills, because it means we get out of our comfort zone all the more frequently.
Surveys give snapshots and comparable metrics; AI chats give narrative detail between cycles. Run both: keep surveys for baselines, use chats for emergent issues and phrasing to mirror back to the org. Less fatigue, richer insight, faster course-correction.
