By combining deep CRM integration and automated WhatsApp claims workflows, Acorn Insurance scaled to 60,000 monthly chats while achieving a 96% containment rate.
Acorn Insurance is one of the UK’s leading specialist motor insurers, offering cover for car, taxi, van, motor trade and home insurance. As a non-standard insurer - serving customers who often fall outside the criteria of mainstream providers - Acorn’s customer base tends to generate more contact than average. Questions are more complex, circumstances are more varied and the need for clear, reliable answers is high.
Acorn’s results at a glance…
- 60,000+ chats automated per month (300%+ increase)
- 96% of interactions handled by the AI Agent
- 23.7% increase in CSAT since user authentication
- 10,000+ images and videos collected via WhatsApp per month
Over recent years, the business has grown significantly across all its product lines. But growth at that pace brings its own challenges. Scaling a contact center to match demand is expensive and slow. The team recognized early that digital servicing, done properly, wasn’t just a nice-to-have. It was the only way to grow without compromising the quality of the customer experience.
While the chat channel was already part of the picture, it was underutilized compared to phone support, which remained the primary contact method for customers. And critically, customers had no way to self-serve outside of core operating hours; something needed to be done in order to keep pace with the company’s increasing customer service demands and provide a better brand experience.
Making smart choices from the beginning
Acorn began working with boost.ai with a clear initial goal: build an AI Agent capable of handling the volume of chat enquiries coming in, and extend the service availability beyond the hours of their contact center.
What followed was not a set-it-and-forget-it deployment mentality. From the beginning, Acorn invested meaningfully in the platform, building out intents, iterating on responses and dedicating internal resources to getting it right. Adam Croughan, Personal Lines Operations Director, is direct about why that matters:
“Delivering meaningful innovation demands a willingness to launch, learn, and continuously improve — backed by sustained investment and organisational commitment. That approach has been central to our success, and boost.ai has helped us realise that vision from the start.” — Adam Croughan, Personal Lines Operations Director, Acorn Insurance
The results speak to that commitment. Acorn has grown from handling around 15,000 chats per month to more than 60,000, with the AI Agent now handling approximately 90% of all conversations.
Leveraging WhatsApp into an automated claims workflow
Of all the ways Acorn has pushed the boundaries of what a conversational AI platform can do, nothing illustrates their approach better than their use of outbound WhatsApp messaging - a capability that started as a marketing experiment and evolved into a core operational workflow.
The original idea was straightforward: when a customer took out a new policy, send them an outbound WhatsApp message to make them aware of Acorn’s self-service portal. It was a low-friction way to drive adoption of a channel the business wanted customers to use. But the intent behind the pilot was always broader. WhatsApp’s native image-sharing capability made it an obvious candidate for something more operationally significant - collecting evidence from customers involved in accidents and claims.
Adam explains the thinking: “It’s quite an emotional time for people if they’re involved in an incident. Giving them a way to communicate with us that they’re used to, and that's easy and fast, that was where the thought process was.”
The old method was fragmented and manual. Reaching customers to request photos was time-consuming and inconsistent, and attaching files to the right case had to be done by hand.
The new process, by comparison, was a different experience altogether. A customer reports a claim by phone, and within minutes, an automated outbound WhatsApp message reaches them, guiding them through uploading photos of video of the incident. An API connector then fetches the media link, and the file is automatically attached to their case - ready for the claims team without a single manual step.
Acorn is now collecting more than 10,000 images and videos per month through this process. Danny Campos, Conversational AI Platform Manager, frames the operational impact clearly:
“Because our claims agents aren’t sifting through images and attaching them to files, they can focus on progressing the conversation, speeding up the process. It’s about taking away tasks that don’t offer value, and allowing our people to get into the areas where they can.” — Danny Campos, Conversational AI Platform Manager, Acorn Insurance
The WhatsApp workflow has since expanded further. When a vehicle is written off, and a settlement offer is ready, Acorn now sends that offer directly to the customer via WhatsApp - including a full breakdown of the payout, how it was calculated and what the next steps are. Customers can accept or query the offer in their own time, without needing to be available for a phone call.
For Adam, the value of this approach goes beyond the operational efficiency gains. “It reduces the administration burden on the conversation with the customer,” he says. “It doesn’t remove interactions over the phone, but it frees our workforce to focus more on the empathy of the situation.”
The CRM integration that changed the game
For many basic chatbot deployments, authentication remains an untapped frontier. The easier path is to build for anonymous users - answer FAQs, signpost customers, and handle high volumes of simple inquiries. But without knowing who the customer is, those interactions can only go so far.
Acorn made a deliberate choice to go further. By connecting boost.ai directly to their core CRM system, the platform can now authenticate customers within the chat session and surface live, policy-specific information in real time. The result is a step change in what the AI Agent can actually deliver.
Danny describes the significance of the shift: “The biggest gain is moving away from being generic to being hyper-specific. Customers - especially in insurance - want that liability to be on us. If we confirm they’re covered for something, they need to know that it’s actually true.”
In practice, this means customers can ask whether they have breakdown cover on their current policy, and get a definitive yes or no, along with a prompt to add it if they don’t. They can ask about upcoming payments and receive their actual payment date and amount. They can query their cover level and get information that is specific to them, not a generic explanation of how policies work.
The impact on customer experience has been measurable. Since the integration launched, CSAT scores have improved from 31.4% to 55.1%. Containment has increased to over 96% and, perhaps most tellingly, external redirects - instances where customers are sent elsewhere to get an answer - have dropped by 61%. Customers are resolving their queries in the conversation, not being bounced away from it.
Adam sees the CRM integration as the foundation for something even more ambitious. “The end goal is to complete end-to-end transactions via chat,” he says. “Full servicing capability - not just answering questions, but enabling customers to actually achieve things. That’s where we are heading.”
AI orchestration under the hood
Underpinning all of this is a shift in how the platform itself operates. Acorn has recently moved to a largely generative AI model through the Boost Agent Orchestrator. This is a layer that coordinates specialized sub-agents for different interaction types, routing conversations to the right agent based on context and handing off seamlessly when topics change.
Around 80% of the AI Agent’s conversations are now handled generatively, with some hybrid flows retained where they make sense. In insurance, where customers need to trust that the information they receive is accurate and consistent, that balance is deliberate - generative AI handles the flexibility and natural flow of conversation, while predefined flows ensure precision where it matters most. For Danny and the team, that transition has been approached with characteristic rigor - “we’ve ticked every box,” he says, “and invented new boxes to tick.”
AI as a force multiplier for the bigger picture
Across all of these use cases, one theme runs consistently through how Acorn talks about their deployment: AI is not here to replace people, but to make them better at the parts of their job that matter most.
Adam is thoughtful on this point: “We want to use this technology to help scale and provide solutions for customers outside of core operating hours, and where they choose to use chat. We're a business built on choice for our customers and our employees - it's about freeing both up to have better conversations.”
The phone channel, he notes, still has a place that technology hasn’t replaced, particularly in claims, where empathy and human judgment are irreplaceable. What the platform does is absorb the administrative load that has historically surrounded those conversations, so that when an agent picks up the phone, they can be fully present for the customer in front of them.
And there’s more to come. Acorn operates across multiple brands and customer types, and the ambition is to extend this capability further across the group. The roadmap includes expanding authenticated journeys beyond the portal, deepening CRM integration and continuing to build use cases that would have seemed out of reach just a few years ago.
“In order to scale effectively, conversational AI is something you 100% need to pay attention to. Invest the time, iterate on it and you will get real value.” — Adam Croughan, Personal Lines Operations Director, Acorn Insurance
