12 Months of AI Prospecting Conversations With Australian Real Estate Principals, Here's What We Learned

Over the past twelve months, we’ve had detailed conversations with principals across residential sales, project marketing, property management, and investor-focused agencies, agencies ranging from boutique independents to network offices managing 100+ listings per month.

The conversations weren’t demos. They were operational discussions: what’s broken, what’s been tried, what the team actually does, where the biggest gaps are, and what they’d look like if they were fixed.

Here’s what we’ve learned.

1. The Database Problem Is Universal, But Principals Describe It Differently

Every principal we spoke to has the same core problem: a database full of past interactions and potential, with no reliable system for knowing who in that database matters today.

But they describe it in different ways. Some say it’s a “follow-up problem”, agents aren’t consistently reaching out to warm contacts. Some say it’s a “database hygiene problem”, the CRM is full of dead numbers and stale records. Some say it’s a “prioritisation problem”, agents don’t know who to call, so they call whoever they can remember.

All three descriptions are pointing at the same gap: the gap between holding the data and having the intelligence to act on it.

The language matters for how you approach the conversation with a principal, but the solution is the same regardless of how they frame the problem.

2. Agents Don’t Fail at Prospecting Because They’re Lazy. They Fail Because the System Gives Them Nothing to Work With.

This was one of the most consistent realisations across every conversation.

Principals who described a prospecting problem usually assumed it was a motivation or discipline issue: agents who weren’t cold calling, weren’t following up, weren’t “working the database.” When we dug into the actual workflow, the picture was different.

The agents weren’t avoiding the work. They were sitting in front of a CRM with 2,000 contacts and no useful signal about which ten of those contacts were worth calling this week. In that environment, prospecting becomes a willpower exercise, force yourself to make calls into the void, rather than a leverage activity.

When we asked agents what would change their prospecting output, the most common answer wasn’t “more motivation.” It was “tell me who to call and why.”

3. Principals Are More Concerned About Brand Control Than Most Vendors Expect

The AI-in-real-estate conversation often focuses on capability: what can the system do? How many contacts can it reach? How quickly?

In almost every principal conversation we had, the first substantive concern was different: will this sound like us?

Real estate is a high-trust, personal business. A principal who has built a brand over 15 years in a suburb has a genuine, legitimate concern about putting their agency’s name on a message they didn’t write and can’t control.

This concern is not irrational. Some bulk SMS tools have caused real damage to agency brands by sending messages that sounded generic, made errors with contact names, or sent to contacts who had previously opted out.

The agencies that engage with AI prospecting seriously, not just for a pilot but as an ongoing system, almost universally require a human approval step before any campaign goes out. Not because they don’t trust the technology, but because the messages are going out under their name and they want to stand behind them.

4. Reply Rates Matter More Than Send Volume, and Most Agencies Have This Backwards

Many of the agencies we spoke to were measuring their SMS campaigns by the number of messages sent. “We sent 5,000 last month” was often the first metric offered.

The number that actually matters is replies: how many of those 5,000 contacts responded with information the agency can act on?

A principal who sent 5,000 generic messages and got 15 replies has spent money and agency brand capital on 4,985 contacts who learned nothing about their situation. A principal who sent 500 personalised, contextually relevant messages and got 40 replies has a pipeline of 40 qualified conversations to hand to their agents.

Volume is easy to optimise. Reply rate requires better inputs, better message construction, and a more disciplined approach to contact segmentation. But reply rate is what creates pipeline.

5. The Agencies Getting the Most Value Have the Cleanest CRMs

This was an uncomfortable finding to share with principals, but it was consistent: the agencies that got the strongest results from AI prospecting were the ones with the most structured, well-maintained contact databases.

Not necessarily the largest. Not necessarily the oldest. The cleanest.

Agencies with consistent note-taking, accurate contact segmentation, and documented interaction history gave the AI system much more to work with when building personalised outreach. The output was messages that genuinely referenced the contact’s history, their enquiry, their open-home attendance, their prior conversation, in a way that felt authentic and drove replies.

Agencies with poorly maintained CRMs got less differentiation in the output, lower reply rates, and less useful segmentation of the responses.

The lesson isn’t that a messy CRM prevents AI from being useful. It’s that CRM quality is a multiplier. The better the inputs, the better the outputs. If you’re thinking about investing in AI prospecting, investing a month in CRM hygiene first will materially improve your results.

6. The Biggest Objection Isn’t Cost. It’s “We Tried Something Like This Before and It Didn’t Work.”

This was the most common sales objection we encountered, and the most important to understand.

Almost every principal we spoke to had tried some form of database outreach in the past: a bulk SMS platform, a call centre service, an outsourced prospecting team. Most had disappointing results. Some had actively damaged their brand with a poorly executed campaign.

When they heard “AI prospecting”, the first mental model they reached for was the thing that had already failed them. Generic messages, no reply, wasted cost.

The conversation that unlocked progress was almost always the same: show them the difference between a generic message and a properly personalised one, side by side. Let them read both as though they’re the contact. The difference is visceral and immediate.

Most principals who saw that comparison were willing to discuss what a well-executed version of the thing they’d tried before would look like. That’s a productive conversation.