I've spent years watching companies promise transformation while delivering incremental tweaks. So when Claudia Nicholls, Chief Customer Officer at Simplyhealth, told me they'd reduced complaints from 600 to 40 while handling 2.5 million customers, I knew this wasn't another cautious pilot project story.
This sounded like a masterclass in fixing what's broken before attempting to grow.
At Salesforce's Agentforce World Tour 2025, Nicholls and Dan Eddie, Simplyhealth's Director of Customer Service, shared how a 153-year-old health insurance provider escaped what Eddie calls "a bit of a pickle."
Indeed, prior to Salesforce's involvement, Simplyhealth struggled with legacy systems that didn't talk to each other, claims that took up to 10 days to process, collapsed phone lines, and complaint resolution times that stretched beyond 16 days.
The company's moment of truth (the claims process) was failing spectacularly.
Members submitted claims after visiting the physiotherapist only to wait five to seven days, sometimes 10, for approval and payment. Customer satisfaction plummeted, complaints piled up, and contact volumes surged as frustrated customers called repeatedly for updates.
Eddie detailed the severity of the situation Simplyhealth found itself in.
"If you've got an outstanding complaint, and if you as a customer have submitted a claim, you don't know where you're at, that's going to drive contact and cost into the business, so it's this self-fulfilling disaster that we were in."
The company's purpose (providing access to healthcare for all) couldn't scale on a foundation of poor experience. Nicholls, a Marketer by trade with 10 years at Procter & Gamble, understood a fundamental truth: spending on brand or performance marketing is wasteful if your experience doesn't match your promise.
So how exactly did Simplyhealth manage to turn this around?
The Five-Year Vision
Rather than chase quick wins, Simplyhealth built a five-year transformation roadmap connecting technology directly to purpose. The strategy focused on three pillars:
- Channel mix evolution
- The balance between human and AI agents
- Operational efficiency that could support growth without proportional cost increases
Eddie explained how his team created a transparent year-by-year vision showing exactly what the future would look like:
"Anyone of my team could say, 'Dan, okay, so in year three, you're bringing WhatsApp in, and you've got a channel shift, and AI is doing 30%, what does that mean for me?' and we can describe what that looks like."
The honesty extended to difficult conversations. When team members assessed their "will and skill" for working in a digital-first environment, some chose to leave. Eddie acknowledges this, explaining that there were "certain jobs that [got] replaced. It's just that, as long as you've got the ability and the will to learn, then it doesn't jeopardize your job."
Over three years, Simplyhealth removed 40% of roles while increasing compensation for remaining staff by 40%. The team became multi-skilled across products, channels, and complex value-adding work.
Starting Small, Scaling Smart
The transformation began with email, chosen deliberately because it allowed careful curation of the knowledge base within Salesforce.
The existing knowledge repository was cluttered with product information and irrelevant content. Eddie's team manually cleaned it out, then filled it strategically with 500 previously sent emails reflecting Simplyhealth's tone and brand, covering three frequently asked questions.
Four people received permission to use Agentforce. The first query ("How do I change dentist?") came through, and AI responded in 1.5 seconds with an empathetic, helpful answer. A human checked it, pressed send, and what previously took 10.5 minutes now took 1.5 minutes.
Seven emails we sent in the first week, generatively with a human in the loop. 14 emails the week after, 21 the week after that, Eddie said.
The team returned to their AI governance forum (co-chaired by Eddie and the CTO, including fraud, compliance, and risk stakeholders) to review customer feedback and validate results before scaling.
Today, Simplyhealth sends 2,000 AI-assisted emails weekly across multiple frequently asked questions.
The company also now operates AI shifts during overnight hours, weekends, and holidays when human staff prefer not to work and customers appreciate 24/7 availability.
The Channel Shift Nobody Predicted
Four years ago, voice calls represented 60% of Simplyhealth's channel mix. Today, it's 12%.
This dramatic shift occurred despite the company's customer base skewing older, a demographic often assumed to prefer phone contact.
"Everything is on your phone. So why not healthcare? Why not insurance?" Nicholls asked. WhatsApp launches in Q1 2026, and Eddie envisions customers staying in their preferred channel for the entire journey, from asking questions to submitting claims to completing renewal documents.
The company is testing in-app messaging and exploring voice-through-channel capabilities, where customers could speak to their phone – asking questions such as 'How much have I got left in my physio pot, and when's my renewal date?' – and receive verbal or text responses without opening an app or making a call.
Governance Without Gridlock
Simplyhealth operates as a regulated business, which means its AI governance framework must satisfy board-level risk appetite while enabling innovation. The AI forum assesses every new AI initiative through a risk framework, evaluating whether existing tools already solve the problem and whether ethical standards are met.
"We're a regulated business, so that allows us to operate within the agreed risk appetite that we have agreed with our board," Nicholls explained.
The best team members have freedom to move faster than planned timelines if data supports it, without waiting for formal approval dates.
Weekly reporting provides traffic light indicators across key metrics. When conversational AI resolution rates dropped over two weeks, the team immediately investigated and discovered a technical glitch.
In addition, the company recently deployed Edge Tier's AI quality assurance platform, which analyzes 100% of chats, voice interactions, and emails, flagging issues in real time.