Case Study: Bluecrest Wellness

Bluecrest Wellness is the UK’s largest direct-to-consumer health intelligence company, delivering more than 1.3 million preventative health assessments across a national clinic network. The contact center had grown on a platform first deployed in 2017 — a system that served Bluecrest well at its earlier scale and that the team chose to evolve as customer expectations and AI capability moved on. Working with NiCE’s UK platinum partner SVL Business Solutions, Bluecrest replaced its contact center estate with NiCE CXone in a fixed-deadline migration, went live with NiCE Automated Summary on the first day of operation in a regulated healthcare environment, and turned the program into an analytics-led foundation driving a five-stage agentic AI roadmap.

“It’s been a huge answer rate improvement. A huge after-in-queue time improvement. A huge after-call work time improvement. The systems are night and day. They really are.”

Russell White

Director of AI & Transformation
Bluecrest

01 BEFORE

A contact center built for one era, ready to be built for the next

Bluecrest Wellness is the UK’s largest direct-to-consumer health intelligence company. Its clinics deliver more than 1.3 million preventative health assessments and tens of millions of blood and physical tests, supporting a population-scale model of preventative care. The contact center is the front door to all of it. Roughly 130 advisors handle around 240,000 voice, email, and live chat interactions a year, scheduling assessments, supporting customers through results follow-up, and managing rebookings.

The contact center had been running on a platform Bluecrest first deployed in 2017, an architecture that served the business well at its earlier scale. The technology landscape had moved on substantially in the intervening years, and so had Bluecrest’s ambitions for what the contact center could be. Russell White, Director of AI & Transformation, joined Bluecrest in 2022 with a brief to build reporting around the estate. The first observation was that the platform Bluecrest had grown on did not have the analytical layer the next phase of the business would require. Marketing campaign attribution had been built around individual phone numbers, an approach that worked when the company ran a handful of campaigns and grew increasingly elaborate as the business scaled. Call dispositions captured commercial outcomes but not the broader picture of why customers were calling.

It took a year of careful work to build the reporting layer on top of the existing estate. By the end of that year, the reporting was clean. What it showed was clarifying. The platform Bluecrest had grown on had served its era well. The era ahead would call for something built for what came next.

02 DESIRE TO CHANGE

A decision not to be stagnant for another decade

In 2023, Bluecrest formally extended that brief into a Customer Services Transformation remit and went to market. The team attended industry conferences and CCS expos to understand what other organizations had built. The shortlist included eight platforms across the major CCaaS providers in the market, scored against a thirty-criteria matrix with heavy weighting on AI capability: AI-driven summarization, real-time agent assistance, AI quality monitoring, omnichannel support, CRM integration, and dialer prioritization.

NiCE scored twenty-eight out of thirty, the highest in the matrix. What weighted the decision was the AI feature set and the pace at which it was being extended.

“I wanted something that was going to continuously develop and improve as we developed as an organization. There is no point in just building something and being stagnant.”

Russell White

Director of AI & Transformation
Bluecrest

Bluecrest already had a relationship with SVL Business Solutions, the UK platinum partner, who was supporting the workforce management estate. NiCE’s partnership with SVL was an additional vote in NiCE’s favor. Procurement started in mid-2024. The contract was signed in December. Go-live was set for 11 March 2025, with no extension option.

03 NiCE SOLUTION

Day-one AI in a regulated healthcare environment

The fixed deadline shaped the implementation approach. Bluecrest consolidated more than 750 inbound numbers, rationalized the IVR estate, and migrated voice and email onto NiCE CXone in under three months. Configuration responsibility lived in-house. Bluecrest built configuration capability through internal promotion, partnering closely with NiCE’s UK platinum partner, SVL Business Solutions.

SVL was already part of Bluecrest’s contact center environment through the existing Calabrio workforce management relationship. That continuity carried into the CXone migration. SVL pushed back on design choices the Bluecrest team initially favored, suggested better routes through the platform, and remained closely engaged through every subsequent capability Bluecrest activated. Live chat went through SVL pre-launch testing, with SVL flagging elements the Bluecrest team revised before going live.

The decision that defined the program was a different one. Bluecrest went live with NiCE Automated Summary, the AI-generated post-call documentation feature inside NiCE Copilot for Agents, on the first day of go-live. In a regulated healthcare contact center, where structured notes carry clinical and compliance weight, that decision was unusual. Most contact center migrations stage AI capability behind a stabilization period. Bluecrest activated it from day one.

The reasoning was disciplined. Bluecrest structured a human-in-the-loop model and made the rules of engagement clear to the floor before go-live. Every Automated Summary was advisor-reviewed before being saved to the patient record. A dedicated Microsoft Teams channel, named the NiCE fault issues log, gave every advisor a direct line to the program leads, the team leaders, and SVL. Anytime a summary was wrong, it was flagged in the channel within minutes. The advisors were the testing layer, and they were empowered to act as it.

“I didn’t see the risk in that, because we had the human in the loop checking it. And it’s not something we’d had before. We just made it really clear to the advisors: This is new technology, experimental technology, and we’re relying on you to fix it. And I empowered them to do that.”

Russell White

Director of AI & Transformation
Bluecrest

Analytics that wrote their own business cases

NiCE Interaction Analytics was the second AI capability Bluecrest activated, and it has produced a different category of result. Bluecrest used Analytics to build categories around active business questions, rather than treating it as a passive monitoring layer. Two cases illustrate the operating model.

In April 2025, Bluecrest launched Health Monitor, a subscription product that brings customers back annually for an assessment, providing both ongoing preventative care and recurring revenue. The contact center was trained to mention Health Monitor on every relevant call. Initial sales were quiet, and the floor reported that customers were not interested. The program leads built a category in Interaction Analytics to measure how often advisors were actually mentioning the product. The category showed mentions running at under ten percent of relevant calls. Training was delivered. Mentions rose to around twenty percent, then drifted back. At the end of June, the team told the floor that mentions were being measured directly. That same week, mentions jumped to forty percent. They climbed steadily after that, reaching close to sixty percent by early 2026, supporting more than 2,200 Health Monitor subscriptions sold through the contact center.

The second case ran in the opposite direction. The product team had launched online self-serve rescheduling and committed publicly that it would meaningfully reduce inbound call volume. A category in Interaction Analytics tracked the figure live. Online rescheduling adoption did rise. Reschedule call volume to the contact center did not fall. Bluecrest had the data to take a constructive position back into the business conversation, supported by its own conversational record rather than anecdote. “Interaction Analytics gives you insight to make informed decisions,” said White, “and to push back on arguments when people are making assumptions that are incorrect.”

Live chat went live in 2026. NiCE Copilot for Agents capabilities including real-time sentiment, real-time call summary, and journey summary are also in production. Knowledge base integration is the next foundational project, with Bluecrest consolidating disparate sources into a single AI-readable foundation before exposing knowledge-driven copilot suggestions directly to advisors.

04 RESULTS

A faster, cheaper, more capable contact center, and a floor that trusts the AI

The operational metrics moved together and they moved fast. Average after-call work time dropped from three and a half minutes to under two and a half minutes within the first two weeks of Automated Summary running on the floor. Answer rate improved from approximately seventy-seven percent in early 2025 to ninety-four percent in early 2026, supported by the new callback capability that holds a caller’s position in queue rather than dropping them into a separate, lower-priority dial list. Average wait time runs at around 137 seconds, with a roughly thirty-second reduction over the prior baseline.

Customer-facing metrics moved alongside the operational ones. Customer NPS rose from 54 to 64 across the period. First contact resolution improved by fourteen percent. Redress, the cost of complaint resolution, dropped by nineteen percent. Trustpilot scores moved from 4.4 to 4.7 over the same window. The post-go-live platform stability profile has been clean.

The financial picture surprised the business in the right direction. The actual landing figure represented a roughly twenty-four percent reduction against the prior year’s spend on the legacy stack. The unspent capital was redeployed into capability the program had not originally committed to fund.

Workforce was reshaped, not reduced. Eleven percent of contact center capacity has been reallocated, achieved through natural attrition rather than redundancy. Some of that capacity supported the configuration and AI roles inside the team. One prior resource planner was promoted to resource planning manager. Real-time analyst capacity was redirected to higher-value work, made possible by NiCE Copilot for Supervisors alerting on threshold conditions automatically. Advisor confidence, measured internally, has roughly doubled. The contact center engagement score has risen by thirty-one points across the period.

Bluecrest’s program leadership has been promoted twice across the program, with AI elevated to its own leadership remit and platform ownership consolidated in-house. The business has been nominated for multiple industry awards across the period. Most consequentially, Bluecrest is using its own conversational data, generated by NiCE Interaction Analytics and TopicAI, to write the business case for its next phase of automation. The data has gone to leadership and ownership. The analytics layer Bluecrest deployed to understand its own operation has gone on to write the case for what comes next.

“Automated Summary for us was a real clear quick win. Even if it takes the advisor twenty seconds to take out a spelling mistake or remove a line that says a customer is vulnerable when they are not, that is still a considerable saving over the five minutes it can take to type up call notes.”

Russell White

Director of AI & Transformation
Bluecrest

05 FUTURE

A five-stage agentic roadmap, built on conversational evidence

Bluecrest’s forward roadmap is sequenced into five stages of agentic deployment, all powered by NiCE AI Agents (Cognigy) and built on the automation opportunities Interaction Analytics has already quantified. Stages one through four work through the foundational customer-facing capabilities: identification and verification at the IVR front door (currently consuming one to three minutes of every existing-customer call), intelligent routing to the right team once verified, automated rescheduling against Bluecrest’s existing self-serve portal, and password resets and knowledge base troubleshooting once the knowledge consolidation project is complete.

Stage five is the one Bluecrest is most excited about. The agentic layer will proactively reach customers who have not returned for an assessment in three or more years, with bespoke return offers calibrated to the customer’s history. For a preventative health business that genuinely believes in annual screening as a health intervention, the ability to systematically bring lapsed customers back is both a commercial and a clinical outcome.

Bluecrest is also evaluating the next wave of NiCE Copilot capability as new functionality reaches general availability. The team is comfortable being a measured early adopter where the gain is clear and waiting where capability is still being filled out. “Some things aren’t always an ROI,” White said. “Some things are a customer experience improvement that has a cost. If you want to do those customer experience improvements and hit the cost case, you need to get your ROI from somewhere else first.”