Simba Sleep is a UK-based sleep technology company specializing in advanced, award-winning mattresses and sleep products. Combining innovative technology with sustainable practices, Simba offers a range of mattresses, pillows, duvets, and bedding designed to deliver the perfect night’s sleep. With over 325,000 five-star reviews and multiple industry awards, Simba is committed to engineering sleep solutions that balance comfort, support, and eco-consciousness.

As Simba Sleep grew from a start-up to a £130 million powerhouse in the competitive world of sleep tech, the challenge was clear: how do you keep customer experience personal and high-quality without ballooning the headcount of your support team?

the challenge: scaling support without scaling support

For Graham Paddon, Head of Customer Experience at Simba, the solution was not just about scaling customer service—it was about doing it smarter. “The smaller you are as a team and a unit, the better you can be. You're nimble, fast, and you hold onto and look after great talent.”

Simba's 24 person customer service team handles all customer touchpoints, including web chat, email, and social community management. But as they expanded into new markets and demand soared, they knew something had to change. The challenge? Maintain the same level of care their customers had come to expect, without duplicating headcount.

Graham’s mission was to manage the explosive growth across multiple markets while avoiding what he called “the compounding costs of scale”—in addition to hiring more agents, team expansions incur additional costs like supervisors, training, and support infrastructure, all of which don’t directly or immediately benefit the customer.

The risk wasn’t just budgetary—it was cultural. A bloated team could make it harder to maintain the agility, talent retention, and customer-first mindset that Simba prides itself on.

future-proofing with generative AI

Simba initially deployed the declarative (or scripted) version of Ada, then briefly tried another platform. But it wasn't until they returned to Ada, this time embracing the generative AI version, that they saw transformational results.

Transitioning to Ada’s AI agent (lovingly named Luna) was surprisingly simple. “It's basically a couple of clicks: point it to its brain, tell it this is your knowledge, finesse it with some guidance, and train it as you go,” Graham said.

Luna now resolves an average of 1,000 conversations per week across channels—that’s the equivalent workload of 8 full-time agents—without increasing headcount. “Equally it's providing 24/7 cover, which if you think about if we wanted to serve customers 24/7, we’d probably double our cost and double our headcount,” Graham added.

With Luna onboard, expanding customer support in lockstep with new markets is no longer a resource-heavy hurdle. Instead, the team can scale customer support instantly. “Whenever we want to move to a new market, it’s great knowing that we can make a few simple adjustments to Ada and that can be the full support function for the first six months or even the first year,” Graham explained. “Just Luna and one AI Manager overseeing her to make sure she’s accurate, on-brand, and delivering high-quality resolutions.”

driving revenue by automating with intention

Simba takes a strategic approach to automation, instead of simply aiming for the highest possible automated resolution rate. They focus on creating extraordinary experiences, leveraging the strengths of automation and personal support to truly add value for both the customers and the business.

Similarly, Luna could easily handle sales inquiries from customers, but the Simba team wants to serve them personally and reach out proactively. They set up guidance for Luna to provide a Calendly link to any sales opportunity or product inquiry so customers can book a callback with the sales team. Luna then enables the sales agents with all the information about the interaction.

This deliberate balance has freed up Simba’s agents to focus on emotive, complex, and revenue-generating interactions, while Luna seamlessly handles everything else.

The biggest win? Reallocating freed-up agent capacity to revenue-generating tasks and processes. By shifting just 3 agents to focus on abandoned carts and sales callbacks, Simba’s customer service team now generates ~£600,000 per month in additional revenue.

“It wasn’t about cutting costs. It was about freeing up the team to do more of what matters for customers—and it’s paying off,” Graham remarked. “We’re doing more for our customers and the business—Luna made that possible.”

Luna: not a tool, but a teammate

One of the reasons Luna has been successful at Simba is because they treat her like part of the team. Rather than thinking about the AI agent as a tool, Graham thinks of her as an agent with unique skills that compliment those of the rest of the team.

Simba set up a quality assurance program with a scorecard that tracks performance according to various parameters that the team needs to meet. Luna undergoes weekly QA meetings alongside human agents, complete with scorecards assessing accuracy, tone, compliance, and customer satisfaction. She even gets surveyed by customers, and her results are benchmarked against the entire team.

This tight integration gives Graham’s team full confidence in Luna’s performance, and ensures customers get consistent support that’s fast, accurate, and on-brand.

Looking ahead

Ada didn’t just help Simba scale support—it helped them scale impact. But Graham and his team aren’t stopping here.

In the spirit of continuous improvement, Simba is already working on expanding Luna's proactive presence across the website, enhancing her ability to take action on customer requests, and continuing to evolve QA and knowledge management practices to keep service sharp.

For leaders considering AI, Graham’s advice is simple.

Head over to Simba Sleep's website to chat with Luna and get a taste of their customer experience firsthand.