On the morning of 16 May 2016, a Christchurch agent arrived at an earthquake-damaged property on Avondale Road. The vendor’s fiancée had invited her over to talk about selling it.

The agent brought a man with her and introduced him as an “EQC specialist.”

He wasn’t. He was a buyer from her private list of “as is” purchasers — investors who bought quake-damaged houses to repair and flip. He’d bought through her before.

He walked the interior. He was in the room when the complainant said she hoped for around $200,000, and when the agent replied that was $20,000 to $50,000 too high.

Twenty minutes after leaving, the agent rang: she and the “specialist” had talked it over, the realistic range was $150,000–$180,000, and — in the complainant’s account — we can get an offer to you this afternoon.

Here’s the detail I keep coming back to. The complaint email landed in the franchise head office’s inbox at 2:21pm.

Three hours and eleven minutes after the agent arrived at her door.

That afternoon the agent emailed an apology and explained the man was a new agent at the franchise, working with her. A lie. During the internal investigation, she named a colleague as the person who’d attended. Another lie. Two days later, she admitted both.

Case one: the warning that changed nothing

Four months later, September 2016, the Real Estate Authority sent her a compliance letter about a separate complaint — discussing a property’s price with a buyer, no agency agreement in place. The letter reminded her of rule 9.6: no marketing without an agency agreement.

It also said something remarkable: the letter would not be recorded on the public register. It would sit on her file, visible only to the regulator, relevant only “if another complaint were received.”

Within six to thirteen days of receiving it, she was emailing her “as is” list again. Over the next five months she sent out details of eight properties — addresses, comparative market appraisals, vendors’ price expectations. Agency agreements in place: zero.

The combined charges reached the Disciplinary Tribunal as [2019] NZREADT 13. Her defence: this is how the industry works — test interest first, paper it later. One earlier CAC decision had even said as much.

The Tribunal demolished that, and set two markers that still matter:

The outcome:

The record between the cases

The 2019 decision wasn’t an isolated event. Laid out flat, the file looks like this:

Four findings in eight years. Remember this list — it’s the whole point.

Case two: the trust account

In October 2024 she was before the Tribunal a third time ([2024] NZREADT 40), now a sole trader under a new trading name.

The numbers do the talking:

The Tribunal was scrupulously fair about what this was not: no theft, no dishonesty, no client lost a cent — every payee was paid on time, because she kept topping the account up.

But it was deliberate. She was running a proper external trust account for other transactions at the same time, so the claim of naivety didn’t survive. Disgraceful conduct, admitted. Censure, half the prosecution’s costs, and a five-year prohibition on any agency employing her — largely symbolic, since the Authority had already declined to renew her licence in 2022.

So — are the two cases connected?

Legally, no. Different committees, different conduct, different years. The 2019 case is about acting without authority; the 2024 case is about client money. Neither charge relies on the other.

Practically, completely. The 2024 panel reconstructed the entire history — 2017, 2019, 2020 — as previous disciplinary history, called it “a highly aggravating factor,” and adopted the Committee’s phrase: a repeated pattern of indifference towards compliance. The prior record is a large part of why the penalty was a five-year prohibition rather than a fine.

In other words: the connection between the two cases is the file. The pattern only exists as a pattern because someone kept the record and could read it end to end.

And notice who that someone was. A tribunal. In 2024. With statutory access to the regulator’s archive. Eight years after the first documented incident.

The memory problem

Now walk through what everyone else could see, at the moment it mattered:

Every new client started her file from zero. The regulator’s memory was complete, but it activates on complaint, investigates for years, and publishes at the end.

Now flip it. The one moment a consumer won an evidential contest in this entire saga was Avondale Road — and she won on a timestamp. The Tribunal preferred the complainant’s account of the phone call precisely because it was “consistent from the time she first made a complaint,” three hours after the visit, in writing. The agent’s story shifted. Contemporaneous record beat reconstructed memory. It almost always does.

The complainant did that manually, on instinct, in an afternoon. Most people never do.

Why I’m building what I’m building

This is the gap Resaido exists for. Not faster agents, not slicker CRMs — everyone is building those. A memory layer: what an agent represented, on which channel, on what date, held somewhere that doesn’t expire after three years or sit sealed in a regulator’s file. Contradictions surfaced when they happen — not in a decision nobody reads until it’s over.

This agent’s file is, in a sense, the product demo. It’s exactly the evidence chain Resaido is designed to produce — except it took two Complaints Assessment Committees, three Tribunal panels and eight years to assemble, and it became legible only after the licence was already gone.

The regulator remembered, eventually. The market never got the chance.


Sources: CAC 414 & 416 v Tafilipepe [2019] NZREADT 13 (liability; penalty at [2019] NZREADT 37); CAC 2204 v Tafilipepe [2024] NZREADT 40; Hammond v REAA (CAC 520) [2020] NZREADT 49; Complaint No C15453 (2017). All findings as recorded in the published decisions.

Building in public - Resaido. Field Notes is where I read NZ real estate disciplinary decisions so you don’t have to.