How Craveable Brands Is Cooking Up an AI-Powered Future

Sam Bradon

Simon Revelman

This article is based on a recent CTO Consulting podcast, Craving Productivity in AI Applications, featuring Sam Bradon (director of platforms) and Simon Revelman, Chief Information Officer at Craveable Brands.

If you’ve never heard of Craveable Brands, you’ve almost certainly eaten at one of its restaurants. The franchisor behind Red Rooster, Oporto, Chicken Treat and Chargrill Charlie’s operates around 620 stores across Australia. Like every quick-service restaurant business right now, it’s under pressure. Cost-of-living headwinds mean customers visit less often and spend more cautiously, pushing the company to look hard at where artificial intelligence can move the needle.

In a recent episode of Inside AI, host Sam Bradon sat down with Simon Revelman, Chief Information Officer at Craveable Brands, to talk through exactly how AI is showing up across the business, from the kitchen to the boardroom and from franchisee support to the drive-thru window.

Winning the Battle for ‘Share of Wallet’

Revelman’s starting point is blunt: in a tight market, the goal is maximising share of wallet, what customers choose to spend at Craveable’s brands versus everywhere else. That means using data and AI to spot trends faster and make quicker calls on product, pricing, promotions and loyalty, while also squeezing efficiency from less glamorous levers like wastage, labour and inventory. None of these moves is dramatic on its own, but stacked together, they add up.

Loyalty, Reimagined

Loyalty customers are disproportionately valuable, visiting more often and spending more each time, so Craveable is rethinking what loyalty actually means beyond the standard ‘spend and earn’ model. Merchandise, event access, and tiered recognition (think gold, silver, and platinum) are all on the table. But the real shift is using AI to segment customers and tailor loyalty offers to individual cohorts, rather than running a one-size-fits-all program.

Revelman doesn’t frame this as reinventing marketing. It’s about speed. Work that once required a data analyst to build customer segments and a marketing team to craft campaigns around them can now happen far faster, with AI helping generate the emails, offers, and creative needed to reach each cohort with something relevant. The strategy hasn’t changed; the cycle time has collapsed.

Franchisees Are Customers Too

As a franchisor, Craveable’s real customers are its franchise partners, and their success is directly tied to the company’s own. On the technology side, that’s translating into automated rostering informed by historical and future sales data (already delivering a couple of percentage points in labour savings), smarter inventory reordering and plans for AI-driven ‘coaching’ built into financial reporting, flagging, for instance, when a store is over-indexing on chicken costs or under-resourced on staff, and recommending fixes.

Because the business runs 24/7 and is genuinely at the mercy of variables like weather and real-time promotions, the ability to react quickly matters enormously, and that’s exactly the kind of anticipatory, always-on coaching Craveable is now trialling.

Choosing AI Vendors in an Era Where Everyone’s Selling AI

One of the more candid parts of the conversation covers how AI is reshaping vendor relationships. Historically, Revelman picked suppliers based on relationships, gut feel and reference checks. Now, with software vendors, consultancies and integrators all chasing AI sales targets, he’s shifted his criteria: he wants proof. That means unpaid proof-of-concept trials, rewards tied to realised outcomes or short free-trial periods rather than multi-year lock-in, a pragmatic hedge against vendors who ‘talk the talk’ before the technology has actually been proven out.

From Handwritten Notes to AI-Scored Interviews

Perhaps the most concrete example in the whole conversation is Craveable’s recruitment process for franchisees, a genuinely competitive market where speed matters. What used to be phone interviews with handwritten notes is now recorded via Teams, automatically transcribed and fed into a custom ChatGPT build that populates interview answers regardless of the order they were given, and generates a sentiment score benchmarked against historical interviews.

The payoff is twofold: recruiters can go deeper in interviews without worrying about note-taking, and candidates move through the pipeline faster, with more consistent decision-making across the whole recruiting team, not just individual interviewers.

The Unsexy Foundation: Data Consistency

Ask any CIO about AI, and eventually you’ll land here: none of it works without clean, consistent data. Revelman is emphatic that a shared data lake with agreed definitions is essential. Gross profit means the same thing everywhere, sales figures reconcile across systems, and customer records are consistent. Without it, any AI agent (or human, for that matter) is left guessing, and every AI-generated report becomes something you have to double-check rather than trust.

Picking the Next Use Case Democratically

Rather than dictating AI priorities from the top, Revelman runs a bottom-up process: train the team on what AI actually is and where it’s suited, then let people identify use cases from their own day-to-day work, such as manual reconciliation, versus something like ad creative, which the group agreed wasn’t a strong fit. The group sets priorities jointly, and critically, AI initiatives aren’t run as a separate innovation stream. They’re folded into the normal project list and re-evaluated alongside everything else, avoiding the classic trap of a shiny AI ‘science experiment’ that never finds a home in the business.

Playing Catch-Up on Training, on Purpose

Revelman is refreshingly honest that staff AI education has lagged. Tools like Copilot went out broadly, some people asked for Claude or Gemini, and usage has been uneven, with licences going underused. The fix: lunchtime learning sessions rolling out shortly, teaching people what each tool can actually do, effective prompting and what to watch out for, aimed at bringing the whole organisation up to a baseline before layering on more advanced skills.

The Three-Year Vision: Self-Serve Everything

Looking ahead, Revelman’s biggest ambition is a virtual agent that can answer almost any question, for anyone in the business. An early version already exists in-store, answering things like food safety temperatures, loyalty point values or youth employment rules, complete with links back to source documentation, instead of employees hunting through policy manuals.

Layered on top of that is proactive coaching for franchisees: alerts that a roster looks too light ahead of a promotion, that stock is running low or that an item needs to be cleared before it expires. These recommendations are grounded in real data, though Revelman stresses that franchisees still make the decisions. Accuracy is non-negotiable here; nobody wants a ‘coach’ hallucinating advice to a small business owner.

Cameras, Kiosks and the Future of Quick-Service

Zooming out to the industry more broadly, Revelman sees several plausible near-term shifts:

• Anonymous personalisation at kiosks, using cameras (without identifying individuals) to infer whether an older customer, a younger customer or a family is ordering, and adjusting the menu display accordingly.

• Personalised ordering for logged-in loyalty customers, including one-tap reordering of usual favourites, since speed is the whole point of quick-service dining.

• Precisely timed loyalty offers, an email at 5 pm on a Tuesday, say, for a customer who reliably visits at 6 pm that day, aimed squarely at beating the competition to that decision moment.

AI at the Drive-Thru: Already in the Lab

Asked directly about AI-driven drive-thru ordering, a technology that’s had a rocky reputation elsewhere, Revelman reveals Craveable is already testing it with a UK-based team. Unlike some market attempts that push upsells regardless of what’s already in the basket, this system is basket-aware, and it’s integrated directly into the point-of-sale system, sending orders straight to the kitchen screen. The plan is to run it alongside a human operator within the next several months, testing it against real accents, real regional slang and real customer quirks, such as ‘Sprite’ versus ‘lemonade’ or ‘Pepsi’ versus ‘Coke’, before considering a fully autonomous pilot.

The Bottom Line

Revelman’s closing take is measured rather than hyped: he doesn’t see AI as a mass job replacer, at least in the near-to-medium term, but as a productivity partner that frees people up for higher-value work. He predicts that fluency with AI will become a standard interview question across every department, not just IT, because no industry or role will be untouched by it.

For a business built on razor-thin margins, high volume and round-the-clock operations, that pragmatic, data-first, proof-before-promises approach to AI looks less like a moonshot and more like the new cost of doing business in hospitality.