Reader launch · $249 for $49
Exclusive Broker Times price Instant download 30-day keep-the-files refund

The AI Annual Review Engine · For every major broker CRM

Broker Times reader launch

The money is already in the book you settled.

$249 $49 AUD · once Save $200
Get the AI Annual Review Engine — $49 AUD$249 · one-time · instant download

They already bank with you. The annual reviews are sitting there unbooked. $49 — not $249 — gets the AI Annual Review Engine: the AI writes a personal review from the settled file, the AI reads the replies, and you only see HOT and BOOKED.

Make.comMake ZapierZapier 12 CRMs
Yours in 60 seconds 30-day keep-the-files refund From The Broker Times, read by 22,000+ brokers
0Make blueprints you
import, not build
0Channels: email,
SMS and AI phone
0Broker CRMs
covered
0minFrom download
to first send
Abstract editorial illustration of an annual review: a calendar marked review due, an SOA-ready summary, an SMS booking, and two faceless silhouettes across a desk

The conversation you are not having, with people who already settled with you.

Built for real broker teams

Every retained client starts as a review you almost let slide.

One broker or a team of nine, the AI works the same way. It reads the settled file, writes a personal annual review, reads the replies, and only hands you HOT and BOOKED.

You are not buying another dashboard that watches the book. You are buying an AI that writes the review, reads the inbox, and optional AI phone nudge — so the cycle actually runs.

  • They already bank with you. These are settled clients, not a purchased list.
  • They already consented. Every message carries an opt-out, honoured instantly.
  • The AI only hands you HOT and BOOKED. LATER and OPT OUT never reach your afternoon.

The problem

The reviews are sitting in the book.

Hundreds of people you already settled. Twelve months on. No review booked, no file note, and the last thing they heard from you was a settlement-day text.

You already have permission to contact them. Some have remortgaged with somebody who did bother to call. Some would stay if you asked. You do not know which, because reviewing four hundred files by hand is nobody’s idea of a Tuesday.

So the anniversary dates scroll past, trail quietly thins, and the best-interest file is thinner than your licensee would like.

The uncomfortable part. A settled book is not a finished job. Every month a review-due client hears nothing, the relationship gets cheaper for someone else to take.

Do the sum first

How many reviews are sitting there?

Conservative defaults. Change the numbers. Nothing is uploaded — this runs in your browser.

Sitting in the book

0 reviews due
0messages that go out
0replies to classify
0clients handed to you
0you never have to chase

The AI writes every review, reads every reply, and puts a CRM note on the record. You only see HOT and BOOKED.

What actually goes out

Not a mail merge. A review written for that client.

The AI reads the settled file — name, suburb, settlement date, loan type — and writes a personal annual review from that. Here is genuine-style output, unedited.

Step 1 · Email
To Priya N. · Ashfield purchase · settled 12 months ago

Hi Priya,

It has been twelve months since we settled the Ashfield purchase, and I wanted to put a proper annual review in the diary rather than let it drift.

Rates, servicing rules and your situation have all had a year to move. A short review lets us check the loan is still doing the job, and I will write up a summary you can keep on file.

Have you got twenty minutes next week?

Matt

Every email carries a one-line, plain-English opt-out.
Hi Priya, Matt from Northside Finance. Your Ashfield loan is due its annual review. Happy to walk through it this week. Reply STOP to opt out.
Yes please. Thursday afternoon suits.
Perfect. I will send two times and put a note on the file.

Only sent when a mobile is on file. The reply is classified as BOOKED.

Step 3 · Optional AI call
AI
Hi, is that Jordan? I am calling on behalf of Northside Finance about the annual review for the loan we settled last year.
Client
Oh right. I keep meaning to book that in.
AI
Completely understand. Would you like Matt to hold a time this week, or should we leave it for now?
Client
This week is fine. Wednesday morning.

Booked · CRM note written

Never quotes rates, never recommends refinance, never gives advice.

The half everyone forgets

Sending is easy. Reading two hundred replies is the job.

A second AI workflow catches every response, classifies HOT / BOOKED / LATER / OPT OUT, and only pings you for HOT and BOOKED. That is the half that turns a review campaign into a system.

HOT

They want the review now. You get pinged. A CRM note is already on the record.

BOOKED

A time is agreed. Deal or task created, summary started, you just turn up.

LATER

Not this month. Parked, tbt_review_due_on pushed forward, picked up again quietly.

OPT OUT

Suppressed immediately and permanently. tbt_review_state = opt_out. No second chances.

One client, end to end

Seven seconds of work. None of it yours.

This is what happens to a single review-due client from the moment the engine picks them up.

Picked upReview-due, consent, frequency cap
AI writes itA personal review from the settled file
SentEmail, then SMS, then the AI call if you want it
AI reads itHOT / BOOKED / LATER / OPT OUT, in about a second
Handed overNote written, you get pinged only if it is worth a call

Multiply that by the settled book, running quietly in the background, while you get on with the files already on your desk.

The install

You import a file. You do not build this.

Three JSON blueprints come in the pack. Drop them into Make and the AI engine appears on the canvas, already wired — every module, every safety filter, every prompt, every field mapping. You set one thing: your own definition of review-due. These are the imported scenarios on the canvas.

Blueprint 01 · the outbound engine
The AI Annual Review Engine outbound scenario on the Make canvas: pull review-due clients, the AI writes the review, then email, SMS and Retell AI phone

Nine modules, already connected. Pull review-due clients, gate on consent and frequency, the AI writes the personalised review, split into email, SMS and the optional AI phone nudge, then stamp the CRM so nobody gets hit twice.

Blueprint 02 · the reply handler
The AI Annual Review Engine reply triage scenario on the Make canvas: a reply comes in, the AI classifies HOT, BOOKED, LATER or OPT OUT

The AI reads every reply. HOT and BOOKED get a task and a ping. LATER is parked 90 days. OPT OUT is suppressed immediately. You never sort the inbox by hand.

  1. Scenarios, then Create a new scenario
  2. The ... menu on the empty canvas, then Import Blueprint
  3. Choose the file. The whole thing appears.
  4. Sign in to your accounts. Done.

About four minutes per scenario. On Zapier there is no import file — Zapier does not allow it — so the Zapier guide walks you through the identical build screen by screen. Longer, but nothing is left to guess. Blueprint 03 is the Sheets bridge for ActivePipe, BrokerEngine and AFG Suite360.

Open the box

This is what you are actually getting.

Not a course. Not a PDF of ideas. The working files, the finished scenarios and every word the system will ever say — already written, already tested, already wired up.

File 01 · Make blueprintImport
The outbound AI review engine on the Make canvas
Nine modules, already connected. The AI writes the review. You upload the file. You do not drag a single module.
File 02 · Reply triageImport
The reply triage workflow on the Make canvas
The AI reads every reply. HOT, BOOKED, LATER, OPT OUT — you only see the ones worth a call.
File 03 · Sheets bridgeImport
03-TBT-Annual-Review-Engine
-SHEETS-BRIDGE.blueprint.json

For ActivePipe, BrokerEngine
and AFG Suite360.
Export the review-due list,
the engine reads the Sheet.
Everything downstream is identical.
The closed CRMs, solved. No aggregator permission required.
The email promptWritten
Write one short annual-review
email to a settled client.

Use: first name, suburb,
    months since settlement.

Rules:
• Under 120 words, plain English.
Never quote a rate.
Never recommend refinance.
Never promise a result.
• One question at the end.
The prompt, not a template. Fresh for every client instead of a form letter.
The prompt libraryWritten
  • Annual-review email — three tone variants
  • SMS follow-up, opt-out line built in
  • AI phone-call script and guardrails
  • Reply classifier — hot, booked, later, opt out
  • CRM note writer
  • SOA-ready review summary
Plain text files. Tune the voice without touching the automation.
The field mappingsWired
tbt_review_state  → due / sent / hot /
                      booked / later / opt_out
tbt_last_review_at → frequency cap
tbt_review_due_on → who is eligible today
firstname → first_name
settlementdate → review_clock
The boring part, done. Create the three properties before you import.
SOA-ready summaryWritten
Draft a file note a licensed
broker can turn into an SOA.

Include: facts on file,
what the client said,
what was not recommended,
next action, date.

It does not write the SOA.
It drafts the note.
In the pack, not an upsell. The review conversation becomes a file you can keep.
The compliance packIncluded
  • Consent gate for an existing-client book
  • Opt-out wording for email and SMS
  • Sending hours and daily volume caps
  • Spam Act, Privacy Act, Do Not Call
  • What the AI is forbidden from saying
  • Pre-launch checklist you tick before day one
Four minutes to read. Existing-client is a stronger position. It is not a free pass.

Pay someone to build it

$2,000–$6,000

A two-workflow build with AI prompts, CRM mapping and a review-summary prompt is a 15–25 hour job at Australian consultant rates.

Build it yourself

15–25 hours

Most of it is not the wiring. It is writing the messages, testing the prompts, and getting the compliance wording right.

Or start from this

$49 · ~25 min

Import three files, connect your accounts, create three CRM properties, test it on yourself. That is the job.

Those rates and hours are indicative, not a quote. The expensive part of this was never the software.

Compatibility

It runs on the CRM you already have

The pack includes a scenario for each major broker CRM. Only the first step and the last step touch your CRM. Everything in between is identical no matter what you run.

SalesforceNative
HubSpotNative
PipedriveNative
Zoho CRMNative
SalestrekkerAPI
ConnectiveAPI
LMG MyCRMAPI
EffiAPI
Finsure InfynityAPI
ActivePipeBridge
BrokerEngineBridge
AFG Suite360Bridge
Native  sign in and go API  subject to your aggregator granting access Bridge  via a Google Sheets staging layer

Straight answer about the closed ones. Some aggregator-owned platforms will not let an automation read them directly, and that access is granted by the aggregator, not by you. For those, the pack ships the bridge method: you export your review-due clients to a Google Sheet and the engine reads that instead. Everything downstream is identical, and it works everywhere without needing anybody’s permission.

What lands in your inbox

One ZIP. No waiting. Nothing else to buy.

A complete do-it-yourself pack for the settled book. Bookmark the download page — updates land there for free.

  • Three Make.com blueprints — outbound engine, reply triage, and the Sheets bridge. Import the JSON. No building by hand.
  • The complete Make guide — every step, every field, every value.
  • The complete Zapier guide — screen by screen, because Zapier cannot import a workflow file.
  • Twelve CRM setup notes — Salesforce, HubSpot, Pipedrive, Zoho, Salestrekker, Connective, LMG MyCRM, Effi, Finsure Infynity, ActivePipe, BrokerEngine, AFG Suite360.
  • The prompt library — email, SMS, AI call, classifier, CRM note, and the SOA-ready review summary.
  • The compliance pack — written for an existing-client book, not a dormant-enquiry list.

How the twenty-five minutes goes

  1. Open 00 Start Here first

    A one-page map of the folder, your CRM, and which guide you need.

  2. Create the three CRM properties

    tbt_review_state, tbt_last_review_at and tbt_review_due_on. Do this before you import anything.

  3. Make, or Zapier

    Make users import the blueprints. Zapier users follow the guide — Zapier cannot import a workflow file.

  4. Connect your accounts

    Your CRM (or the Sheet), OpenAI, your mailbox, your SMS provider, and Retell AI if you want the phone channel.

  5. Test it on yourself, then open the tap

    Add yourself as a review-due client, run it once, reply STOP, confirm it suppresses you. Then fifteen clients a day for the first week.

Who made this

Built by the people who write for 22,000 brokers every week.

The Broker Times covers what is actually happening in Australian broking: aggregator platforms, compliance, lender policy, and the tools brokers are quietly using to get their week back.

This engine came out of that. It is built around Australian rules, Australian language, and a CRM that has been filling with settled clients for six years. Not a generic nurture sequence with the word “review” pasted in.

Nobody needs another dashboard. Brokers need the annual review to happen without them chasing it.
Abstract constellation of a broker settled book, with review-due client records lit in coral around a central hub labelled The Book

Runs the same whether you are one broker or a team of nine.

Buy this if

You have a settled book with reviews stacking up, you want the AI to write the reviews and only hand you HOT and BOOKED, and you would rather spend twenty-five minutes once than chase anniversaries by hand forever.

Do not buy this if

You want somebody to build it for you. This is deliberately do-it-yourself, which is the only reason it is $49 and not $2,000. Also skip it if you wanted the Lead Reactivation Engine — that product works dead enquiries. This one works existing clients.

30
DAY

Read it, build it, and if it is not for you, tell us.

Email matt@thebrokertimes.com.au within 30 days of purchase and we will refund you in full. You keep the files. We would rather you had a good experience than $49 of ours you resent, and it is the same guarantee we put on everything else in the shop.

Before you buy

Straight answers

Do I need to be technical?

No, but you do need to be the kind of person who has connected two accounts together before, or is willing to follow a guide carefully for twenty-five minutes. If that is not you, this is not the right product and we would rather tell you now.

What does it cost to run?

Roughly seven Make operations plus a fraction of a cent of OpenAI per client contacted, on top of your own SMS and call charges. For most brokers that lands in the low tens of dollars a month. The clients themselves cost nothing, because you already settled them.

Is this compliant?

The engine is built around an existing-client book. It only contacts people already in your CRM, every message carries an opt-out, opt-outs are honoured immediately and permanently, and the pack includes a checklist covering the Spam Act, the Privacy Act and the Do Not Call Register. Existing-client is a stronger starting position than a cold enquiry list. It is still your database and your licence, so you make the final call.

Will the AI say something stupid to my clients?

The AI is forbidden from quoting rates, giving credit assistance, recommending refinance or stay, promising a result, or manufacturing urgency. Before you go live you run it on yourself and read the output. We tell you to read the first ten reviews it writes to real people, and you should.

Does it work with my CRM?

Yes. The pack includes a setup note for each major broker CRM in the grid above. Native where the CRM allows it, API where your aggregator grants access, and a Google Sheets bridge for ActivePipe, BrokerEngine and AFG Suite360. If yours is not on the list, the bridge is the fallback.

What exactly do I receive, and when?

A ZIP, on screen the moment you pay, and emailed as well. Three Make blueprint files, both setup guides, twelve CRM notes, the prompt library including the SOA-ready review summary, and the compliance pack. Bookmark the download page and you can re-download any time.

Do you install it for me?

No. We have kept it do-it-yourself on purpose so the price stays where it is. Everything you need is in the guides. Support is matt@thebrokertimes.com.au.

Why is it $49 when it says $249?

Because this is the Broker Times reader launch price. $249 is what it is worth against a single retained relationship. $49 is the call for readers of The Broker Times. Buy now and later updates are free. It is not $149 — that is the other engine.

Is this the Lead Reactivation Engine?

No. That product works dead enquiries. The AI Annual Review Engine works the settled book. Different job, different prompts, different CRM properties, different price.

Abstract editorial illustration of an annual-review conversation and an SOA-ready summary on the desk

What it is all for

It ends in a review meeting, not a dashboard.

Every automation on this page exists to produce one thing: a client who already settled with you, picking up and saying yes, actually, let’s go through the loan.

Monday morning. Four replies came in over the weekend. The engine already read them, sorted them and wrote the CRM notes, so the only decision left is which review you take first. The SOA-ready summary is waiting on the ones you booked.

The book is already there

$49 to start the reviews sitting in the book you already settled.

The AI writes the reviews, reads the replies, and optional AI phone nudge. Three blueprints, both guides, twelve CRM notes, the prompt library and the compliance pack. One payment, instant download, yours to keep.

$249 $49 AUD · once
Get the AI Annual Review Engine — $49 AUD$249 · save $200 · instant download

Or see how many reviews are sitting there first.