The betting and iGaming industry, explained

Verte News

Edition: English

Betting News

Raphie founders warn operators are automating support for the wrong metric

The co-founders of AI player-support firm Raphie tell SBC News that deflection-based automation hides financial distress and harm signals that support desks are meant to catch.

Illustration of AI-driven customer support automation for a betting operator
AI-generated illustration

Betting operators racing to automate customer support risk optimising for the wrong number, according to the founders of AI player-support firm Raphie.

Justin Heath and Alex Jones, who co-founded Raphie, told SBC News that most automation tools are built to contain and close a conversation rather than resolve it, and that the figure reported back to operators is simply how often that containment happened. A player who abandons a chat in frustration scores the same as one whose issue was actually fixed.

"An abandoned conversation is categorically a negative outcome," Heath said. In a regulated betting market, he argued, it can be worse than that, because the support channel is often where an operator first spots financial distress, confused bonus terms, a stuck withdrawal or early signs of chasing losses.

Heath and Jones built their view from the operations side before building the technology. Prior to founding Raphie, they ran support and payments operations for betting brands, at one stage managing hundreds of agents across multiple operators, drawing on eight years of experience and more than a million player interactions.

"We ran support for operators before we ran AI for them," Jones said. "That order matters. You cannot automate a process well if you have never had to run it during the peak events and address issues specific to the industry."

That background, the pair said, exposed a structural problem with generic customer-service AI: roughly 90% of iGaming contacts require account-specific detail about a particular player, bet or pending withdrawal, which tools trained on generic help-centre content and chat logs are poorly equipped to handle. Heath said operators often discover that a tool marketed at 70% or 80% automation delivers closer to 5% or 10% once it meets live tickets inside systems such as Salesforce or Zendesk.

Raphie's own reporting is built around what it calls true resolution rather than deflection, requiring three conditions: the player rates the exchange as satisfactory, no human agent intervenes, and the same player does not return about the same issue within 24 hours. On that measure, the company says it achieves roughly 74% true resolution and around 88% customer satisfaction across its client base.

"Containment flatters everyone and fixes nothing," Jones said. "Resolution is the number that shows up later in retention."

The model is designed to expand gradually, starting with a narrow scope and widening category by category as it proves itself, with compliance flags, vulnerable-player cases and disputed bets routed to trained specialists rather than handled automatically. Raphie treats that escalation path as core to the product rather than a sign of failure, an area the founders say regulated operators scrutinise closely.

Jones said the handover to a human agent is often seamless from the player's perspective: "We see people wishing Raphie a good night and signing off."

The company cites an OSB operator in Arkansas as a case study, where it set up the human support and payments operation first, then introduced automation gradually. After two months of full automation on a defined set of ticket types, the operator saw an autonomous resolution rate of 77%, satisfaction readings topping 90%, and response times that had fallen by over half.

Across its wider client base, Raphie says it has reduced cost-to-serve by more than 60%, with a typical go-live period of around ten weeks. The founders say the pattern holds across their book of business: operators that get the underlying support operation right before automating it see stronger results than those that automate first.

Raphie is now expanding into European markets, where rising gambling taxes and tighter margins are pushing operators toward automation faster than in the US. The company says it is rebuilding its compliance and escalation layer market by market to align with local player-protection rules and audit requirements, while transferring existing knowledge on payments, bonuses and KYC processes.