FROM LEGACY CHAOS TO CLOUD CLARITY: AN AMBITIOUS MIGRATION PRESSED FOR TIME
If you had asked me two years ago whether we could migrate the contact center voice infrastructure of the Netherlands' largest energy supplier in under 12 weeks, I would have told you it was very ambitious, but doable with the right team. Most of my colleagues at the time told me I was mad.
It worked. And I'd like to share how we did it, what we learned, and where we're heading next.
The starting point: Systems that were holding us back
Customer experience is our primary competitive differentiator at Essent. A large part of our business model is selling & serving commodities which by nature is a very generic business, electricity is electricity and gas is gas. What keeps customers with us or what makes them leave is the quality of their experience every time they interact with us.
That makes our contact center set-up genuinely strategic. Yet by late 2024, the systems underpinning that set-up had quietly become a liability.
Our on-premises legacy voice platform had gone end-of-support. Sabio, our long-term service partner, was helping us keep it afloat. In all honesty it had become an innovation bottleneck. Configuring new routing logic, deploying automation, experimenting with AI. All of it was either painfully slow or simply impossible on the existing stack.
Our live chat solution had its own set of problems. Originally built for e-commerce, the suppliers’ roadmap had diverged from what a utility service provider needs. We'd accumulated custom builds and workarounds over the years, creating technical debt that required ongoing maintenance without delivering real customer value.
This resulted in a sub-par day-to-day experience for our customer agents. Voice and chat lived in separate systems. Agents managing multiple chat conversations would have multiple browser windows open simultaneously.
We had a situation where our agents were constantly working with different screens, and if they had multiple chats, they had multiple agent desktop instances open. This was not only messy but, I can imagine, in some cases quite stressful.
Choosing our path: Genesys cloud and a partner who'd done this before
We didn't start from scratch on the technology decision. One of the advantages of being part of the E.ON Group is access to peer knowledge across European markets. Colleagues at other E.ON entities had already gone through similar contact center transformations and we were able to reuse much of their market research. This saved us significant time.
The feedback across the E.ON group on Genesys Cloud was consistently positive. Sabio also proofed to be a no brainer. They were already embedded as our partner and had over 25 years of Genesys expertise under their belt. This combination gave us real confidence.
The decision to consolidate voice and chat onto a single cloud platform via Genesys Cloud was straightforward once we'd done the homework. The harder decision was how to execute it.
The execution: Phased, Agile, and genuinely embedded
We made two structural choices that I believe made the difference between success and a very stressful failure.
First, we chose a phased approach over a 'big bang' cutover. Rather than switching everything at once, we started with our SME customers first. A distinguished group of customers with dedicated agents and isolated phone lines. This gave us a controlled pilot environment to identify and resolve configuration issues before they could affect the main operation. Our agents got real-world experience with the new platform in a lower-stakes context. By the time we scaled to the full operation, everyone had more confidence.
Second, we adapted our implementation execution model mid transition. We started with a more classical Prince2 project approach. In the first week this approach was to conflicting with our internal way of working. Together with our partner we came up with a new way of working. Sabio's consultants joined our agile sprint teams in 's-Hertogenbosch, working shoulder-to-shoulder with our conversational designers, engineers & customer contact experts. Instead of working from stage gate to stage gate we worked on multiple streams in parallel.
What we really appreciated in the partnership with Sabio is the flexibility they provide. We had to move from a more classical project management approach towards a more agile way of working. The Sabio experts became truly part of our agile teams.
The voice migration was a high-stake game due to the contract with the legacy supplier ending within 3,5 months. We managed to complete the full migration under 12 weeks from project kick-off to go-live. The live chat migration followed six months later, including a deliberate decision to execute the chat cutover in December, right in the middle of our peak season. Not making this deadline would result in a costly extension which would press heavily on our operating expenditure. You can imagine that this choice made some colleagues nervous. More on that below.
Beyond the core migration, we took the opportunity to include quality of life improvements that had previously been impossible. For example, we unified our CRM directly within the Genesys Cloud UI, giving agents a single screen to handle voice and chat alongside customer data & transactions. This not only resulted in efficiency gains but also created buy-in from our end users.
The results: What the numbers say
- 50% reduction in contact center technology costs
- 0 critical incidents since go-live
- 36% improvement in agent workplace satisfaction
The cost reduction confirms what our business case projected: eliminating hosting, legacy licensing, and specialist maintenance overhead delivers real savings. But the number I find most meaningful is the agent satisfaction score.
In our quarterly employee NPS measurement for the agent workplace, we saw a significant uplift. That's not a soft metric. It reflects a genuine change in how our people experience their working day. Less cognitive load, fewer systems to juggle, less risk of errors. Agents can focus on the customer in front of them rather than thinking about which window to open.
And do not forget about the December chat migration. Despite executing a major platform switchover during peak season and pursuing ambitious customer NPS targets, our live chat NPS didn't drop. It actually lifted slightly. The migration was, in effect, invisible to our customers.
We made the switchover and the customer experience remained at the same level. It even lifted up a little bit. The impact that a lot of people feared was on technical issues and agent adaptation challenges did not manifest. We delivered a silent end to end release .
Zero critical incidents since go-live is worth pausing on too. Before migration, P1 and P2 incidents affecting customer access were sadly a regular occurrence. Since moving to Genesys Cloud, we haven't had a single one originate from our live channel management stack. For an energy provider where customers may need to reach us urgently operational stability is essential.
What's next: Building on a stable foundation
The migration was never the end goal, it was a first step. With stable, scalable cloud infrastructure in place, we're now focused on what we actually wanted to do all along.
Looking ahead, the roadmap includes expanding conversation intelligence like call summarization, contact categorization at scale and exploring more sophisticated LLM-powered self-service capabilities for customers. We're also evaluating modern workforce management solutions that give both our operations teams and our agents better visibility and flexibility around scheduling.
Next to that we're looking at bringing our email, social channels & other work onto our “agent desktop”. The consolidation logic is compelling: one place for our agents to handle everything for our customers.
The fundamentals need to be right because you need a great basis to build upon. We're looking at agentic AI and experimenting with it, but you need stable, scalable customer interaction infrastructure first. That's what we now have.
A personal reflection
I'll be direct: there were moments during this project where the timeline felt genuinely risky. Migrating mission-critical infrastructure for millions of customers on strict deadlines, while maintaining service levels and managing a year-end peak. That is not a trivial undertaking.
What made it possible was a combination of a phased implementation strategy, a partner who embedded themselves in our way of working rather than dictating their own, a solid roll out plan within our operation and a willingness to be ambitious.
For other product managers and transformation leads navigating similar decisions: the lesson I'd take from this is that these kinds of migrations don't have to be years-long, high-disruption programs. With the right partner, the right methodology, and a willingness to move decisively, you can deliver significant change quickly without disrupting customer experience.