Crayons CRM cut 47 leasing calls a week, something print shops have seen before

Pure Real Estate, a Brisbane real estate agency, says an AI chatbot now sits in front of leasing inquiries and that incoming leasing calls fell by 47 per week as a result. Its in-house tool, Crayons CRM, recently won the People’s Choice Award at the REACH Australia and New Zealand Top of the Props pitch event, according to Elite Agent.

For enterprise operators, the takeaway is less about the trophy and more about what was tracked. It is a specific “call deflection” result connected to a single intake step: handling routine availability, application, pets, and property-feature questions before a staff member gets involved.

Call deflection is a clearer ROI metric than “AI productivity”

Elite Agent reported that Pure Real Estate’s team estimated it was fielding 47 calls a week from prospective tenants who were largely asking the same questions. That baseline mattered because it indicated the workload was repetitive and frequent enough to consider automation, without relying on hypothetical time-saved or broad “efficiency” claims.

Similar operational reasoning is discussed in other service-heavy sectors. In its August 25, 2026 Technology Outlook programming focused on software, workflow, and front-office automation, WhatTheyThink argued that many print operations spent the last decade automating the production floor while the front office still relied on email chains, manual rekeying, and tribal knowledge. WhatTheyThink added that when customers want quick turnaround and self-service, the constraint shifts to the front office rather than the pressroom.

Taken together, these examples point to a narrower, supportable point: teams make a clearer case for front-office automation when they can report outcomes in operational units, such as calls, touches, or appointments, rather than broad “agent productivity.” This is most relevant for organizations with high inquiry volume, messy intake across channels, or staff already consumed by exception handling.

Crayons CRM uses image analysis and supports offline workflows

Elite Agent’s description of Crayons CRM goes past simple keyword-style Q&A. Co-founder Milo Holmes said the system performs image analysis on listing photos and uses those inferences to answer feature questions, including whether bedrooms are carpeted, whether a wall shows a split-system air conditioner, or whether a cooktop is gas or electric. Operationally, that means teams can respond to “is it suitable for me?” questions without rewriting listing scripts or building a manual knowledge base for each property.

Elite Agent also described a case in which a prospective tenant asked whether an older, brick, top-floor apartment was wheelchair accessible. The chatbot reportedly used context, including no lift being mentioned and the unit being on the top floor of an older building, to indicate stairs were likely and offered escalation to a manager. Operationally, that matters because it can avoid two costs at the same time: a phone call and a low-value inspection that becomes a no-go once the tenant arrives.

Another operational detail is resilience when workflows get messy. Elite Agent reported that Crayons CRM can operate offline to cancel no-show inspections. Practically, that suggests the team designed for edge cases where connectivity or app use fails, a class of last-mile problems that front-office automation projects in other industries often underestimate.

A shared lesson with printing: the buyer sets the service-speed baseline

WhatTheyThink’s Technology Outlook makes the point directly: buyers judge B2B interactions against the best self-serve experiences they get elsewhere, and front offices built around inboxes and rekeying have trouble meeting that bar. Leasing inquiries can behave the same way. If the only route to a basic answer is a phone call, people will call, and staff end up acting as human routers all day.

Elite Agent reported that Pure Real Estate’s founder Russell Peter was paying nearly $900 a month for software while still dealing with heavy inbound call volume. For operators in any vertical, the procurement lesson is familiar: higher spend does not guarantee higher throughput. A stack that mainly pushes links without answering questions can drive more contacts instead of fewer.

The cross-industry takeaway is strongest for organizations with structured “objects” plus analyzable media, such as a property listing with photos, a print job ticket with PDFs and specs, or a service request with images. In those settings, automations that can interpret unstructured inputs, not just route forms, are more likely to reduce the number of touches in the process.

Questions to put in your next front-office automation RFP

• What is the first measurable unit the project will reduce: inbound calls, inbound emails, manual rekeying events, appointment no-shows, or quote-cycle hours? Request a baseline count and a post-deployment reporting method, similar to Pure Real Estate’s “47 calls per week” calculation (Elite Agent).
• Can the system answer questions using existing assets, such as photos, PDFs, or job files, or does it require manual knowledge-base buildout for every item? Crayons CRM’s listing-photo analysis is a concrete reference point for what “no extra data entry” can look like (Elite Agent).
• Where does the workflow go when the automation is uncertain: does it escalate with context, or does it dump the customer back into a phone queue? WhatTheyThink’s front-office outlook emphasizes moving beyond email-thread triage to structured handoffs (WhatTheyThink).
• What happens when systems are offline or channels break: do staff still have a way to prevent wasted appointments or unnecessary production starts? Elite Agent’s note that Crayons CRM can cancel no-show inspections offline is a useful resilience test case (Elite Agent).