What if a member experience became so good it could unlock multi-billion-dollar growth and contribute to a company going public?
It sounds like a leap, but small-scale studies an even an academic paper have linked stronger customer outlooks with stronger company performance, particularly when it comes to an initial public offering (IPO).
Oura Health Oy was founded in 2013 in Finland by Petteri Lahtela, Kari Kivelä, and Markku Koskela. Its revolutionary smart ring provides members with round-the-clock real-time bio signals related to everything from fitness to sleep patterns and even resilience to stress.
The intent is to make health a daily practice for wearers. "That's our mission," says VP of member experience (MX), John Moses.
Oura's first smart ring launched following a Kickstarter campaign in 2015, but it wasn't until recently that things really started to heat up.
In October 2025 – the year Oura surpassed five million ring sales – the company closed its Series E funding round at more than US$900 million, pushing its valuation to an eyewatering $11 billion. Its 2025 revenue is reported to have reached $1 billion, double the 2024 figure, with sales projections for 2026 standing at $1.5 billion at the time of writing.
Growth like this – particularly in a highly competitive and still emerging niche such as smart rings – doesn't happen by accident. And as great as any product can be, it also doesn't happen without a customer-first mindset.
As a specialist in re-engineering customer engagement to drive retention, lifetime value, and brand reputation, Moses joined Oura in 2023. He says that at the time the MX team was "a tiny little start-up doing email-only support with very poor outcomes".
"We saw that we had to do some major work," he adds and it's here where things get interesting.
In transforming its member and service experience, Moses became acutely aware that legacy metrics are insufficient in the AI age and a clear industry gap was identified: CX needs a new metric.
Understanding MX is the source of all growth, Moses embarked on a side quest to find a better way to measure the experiences Oura delivers. He explains: "We saw a big gap in the CX tools and the measurement we had in the AI-first world – measurement needed to evolve alongside the automation we're seeing. Now we're trying to share that blueprint."
In this interview with CX Network Moses talks about Oura's total customer experience score (TCX) and explains how it can solve CX's most pressing challenges and what it all means for the future of customer listening.
What is the total customer experience score?
The total customer experience score (TCX) is a framework built using large language models (LLMs), which utilizes artificial intelligence (AI) to evaluate and analyze nearly every member interaction across every channel.
In gleaning insights into more than five million wearers, Oura can deeply understand MX for close to 100 percent of the customer base, not just the segment that would – before the AI age began – traditionally complete a survey.
"We have a saying at Oura; the best survey is the one you don't have to send. We shouldn't have to ask in this day and age; we know and we can find out based on the signals members give us," Moses says.
"Instead of asking members to rate an experience after the fact, TCX reads the actual conversation and evaluates what the member experienced. And it's doing that across chatbots, live chat, email and phone interactions," he explains.
This isn't simply about efficiency or service quality. Oura says TCX is about strengthening the relationship members have with the brand; an abstract concept which, despite all the new tech to support experience delivery, remained elusive.
How does TCX work?
In short, TCX intentionally ignores internal notes and backend data and focuses on what the member went through by receiving signals throughout the support interaction, across all utilized channels.
Four AI classifiers score every interaction, each answering one question:
- Issue identification: Was the member's issue fully and correctly understood?
- Frustration: Did the member express frustration with the support experience?
- Resolution: Was the issue resolved or properly escalated?
- Satisfaction: Was the member's sentiment positive after the solution?
An interaction only "passes" if all included questions show a positive outcome. For example, a conversation that solves the problem but leaves the member dissatisfied still counts as a bad conversation.
The scores are then combined into a single score that gives teams a shared view of MX.
Oversight remains critical. Moses says every quarter, results are "calibrated against a golden test set of at least 1,000 cases we've rated". In the event of a disparity, human oversight confirms the standard, and the model is re-tuned so its outputs stay benchmarked against that standard.
How does TCX solve the metrics problem in CX?
In short, TCX drills down into the micro-level details that other listening mechanisms miss.
For Oura, lagging CX metrics such as CSat and NPS, don't provide the data to reach a point of statistical validity on sub-segments of the business. "With TCX, we now have that," Moses says.
"We built TCX to provide real time, 100 percent population oversight across both human and AI interactions. That's really important because the QA calibrators weren't doing AI, it was left unchecked and we needed to have a similar model across both. We're then turning that added visibility into what we consider operational speed. We've reduced the time from problem to understanding to action, from weeks to days, if not hours," he adds.
On how TCX achieves this, Moses says there are four fronts:
- TCX confronts the lack of engagement CX has traditionally seen with surveys while also overcoming the fact that humans are "not naturally inclined to rate a bot".
- It's holistic. Did the customer express frustration with their service experience? Was their problem solved, escalated or elevated appropriately? What can be learned from their cues? "It's not just SOP compliance with issue identification. It's not just sentiment and effort with frustration. It's not just resolution, CSat or NPS. All of that gets combined," Moses says.
- In reviewing both AI and human support, TCX is complete and consistent. It's also "totally objective" because a computer is doing the rating Moses says.
- The fourth front is that TCX is the "divining rod". Moses says: "We can see where experiences are breaking down and why, then we can drill into those underlying conversations. And the 'we' here is critical; everyone is in on it. TCX empowers every layer of our support network to see these breakdowns. With CSAT return rates and lags, this was impossible."
Moses adds: "TCX supercharges our whole team's understanding, from me, the VP, down to every member care representative. There's enough data for every slice and every cut of the business. I describe it as a Rubik's cube that I feel like I now have, and I feel empowered!"
Why is TCX relevant now?
The internal story from Oura is that brands have long struggled with how traditional VoC methods skew overall perceptions and satisfaction. The loud minority, the silent majority, the missing middle. It all misrepresents what's actually going on, turning a lack of visibility into a management problem, not just a measurement problem.
For a start-up carving a reputation in a high-stakes and highly competitive industry, an exemplary MX was critical. Moses knew it, the executive team knew it, and the board of directors knew it.
Instinctively, Moses also knew where improvements needed to be made – he'd spent the 30 years before this appointment doing the same for many other brands – but sample sizes and completion rates were amplifying the wrong signals.
"The loud minority was winning and overshadowing the real improvements we were making," Moses recalls. "As a performance evaluator, we didn't have a good signal."
The development of TCX, in Moses' words, was "rigorous", and took around a year.
"We looked at over a dozen possible classifiers or questions we wanted AI to ask. And then we narrowed it down to four highly-reliable questions," he recalls.
Taking a broader industry view, the relevance of TCX comes back to the irony in how AI delivers on its CX promises: AI-powered technologies have the potential to transform service and brand experiences for the better, but AI-powered mistakes cascade with equal velocity. As AI is scaled across businesses, stronger oversight of the customer's experience is required.
"We're at a moment like never seen before. It's radically changing customer service, and it comes down to scale and risk," Moses says, and it means oversight must scale in tandem with the AI. "This metric, I think, gives you that oversight to keep up when scaling AI to do the service work. Now you've got to scale the oversight," he adds.
What does TCX mean for the future of customer listening?
It's a sweeping statement, but it's relatively safe to say every brand on the planet is trying to understand – and embrace – the future of customer listening. In Moses' view, "it's time for the customer service industry to find and adopt new measurement standards".
While some are tapping the data captured by their in-ecosystem conversational AI platforms, others are using agentic twins or synthetic data. TCX is Oura's solution and the principles and logic can be taken as a blueprint and applied to other organizations.
The biggest draw, however, isn't simply that TCX is a metric for the AI era; it's that Oura's story is a testament to the value of granular-level observability.
Today, Moses describes Oura as a membership business – rather than a tech, data or wellness company, although it is those things as well – and it exceeded the five million paid member milestone this year. The Oura Ring 5, the world's smallest smart ring, launched in May and at the time of writing, Bloomberg is reporting that this remarkable transformation and growth is likely to culminate in an IPO as soon as this month.
Oura itself has yet to issue a statement on the IPO, and it doesn't double down on these links itself. However, those in CX know there is always a connection between outstanding and measurable experiences, product quality, record sales and the march towards stronger financial performance, or in this case, a public listing.
As outlined earlier, there's a scientific paper that examines how higher IPO first-day returns and trading volumes – as well as greater gross profitability three years post-IPO – track higher when CX is measurably better. And that was using traditional metrics such as retention, acquisition and lifetime value. Adding weight to the argument, a 2026 study by Watermark Consulting found companies with the best CX indicators outperform the S&P 500 stock index by more than 400 percentage points.
Turning attention to the wider CX space, Moses says: "In 2027, AI observability is going to become as important as the automation itself. We need to be able to see if interactions are actually good.
"Customers are judging agents and saying which companies have bad service. They're going to be really quick to judge the AI from those companies as well, which means we've got to make sure those interactions are good," Moses explains.
He adds: "CX measurement is going to move from sampling to full population analysis, and that's what we're proposing with TCX."