Who they are
The Real Estate Institute of Queensland has been the state’s peak body for real estate since 1918, representing more than 15,000 real estate agents and agencies across residential, commercial and property management. Its member CX team fields everything from membership and training to advisory and events.
The problem: a queue full of email nobody needed to read
As membership grew, so did the noise. A large share of what landed in REIQ’s queue never needed a human at all auto-replies, bounce-backs, system notifications but each one still arrived as a ticket someone had to open, read and close.
The problem wasn’t member demand. It was the volume of low-value email sitting on top of the real work.
The obvious fix was also the risky one. A blunt rule that closed anything looking routine would eventually close something that wasn’t, and losing one real member enquiry costs more than the hours it saves.
What we built: a system that understands the ticket, not just the subject line
Instead of writing rules against keywords, we used Zendesk’s AI topic detection and intelligent triage to read and classify every email on arrival then built the business logic around what it found. Each ticket is resolved, routed to the right team, or held for a person, according to rules mapped to how REIQ actually works.
Three principles shaped the build:
- Watch first, automate second. We shadowed the team through normal days and pulled apart real tickets before writing a single rule.
- Automate the safe, protect the sensitive. Routine traffic clears itself. Anything touching member records stays with a human.
- Act only on confidence. When the AI is sure, it handles the ticket. When it isn’t, a person gets it.
That sat inside a wider re-engineering of REIQ’s Zendesk instance: a simpler agent experience, intake rebuilt so membership, training, advisory and events enquiries land with the team that can answer them first time, and an internal knowledge base so staff including several new starters give consistent answers from day one.
“This is us implementing AI in the business in a considered way, with a trusted partner who already knows the product.”
Kat Beavon, COO, REIQ
The results
The results
380
tickets auto-resolved per week, with no agent touch
3,656
ticket handled by AI in the past 3 months
~312
of agent time back every year
And REIQ can see it working. Live reporting shows exactly what the AI is handling and what type of work it is, so the result isn’t a claim it’s a number the team can check any time.
“The automation took a job off our plate that never should have been a job. It’s handled before anyone opens the queue, and my team can focus on the members who actually need us.”
Ruth Isaac, Membership CX Team Leader, REIQ
Why it matters
For a member organisation, every hour spent closing email that needed no reply is an hour not spent on the people paying for the service.
By understanding how the team really worked, automating only what was safe to automate, and rebuilding the platform around it, we turned a queue that clogged itself into one that clears itself.
That’s how Gravity CX works: find what others miss, automate carefully with the right guardrails, and give support teams their time back for the work only people can do.
