All Case StudiesMortgage Brokers

The First-Home Buyer Who Messaged at 9pm and Got an Answer

How AI captures and qualifies mortgage leads after hours — answering KiwiSaver questions, explaining LVR rules, and booking consultations while you sleep.

4 min readUpdated 2026-03-15Based on Claude Sonnet 4 / GPT-4o

The Real Problem

It's 9:15pm on a Thursday. A 28-year-old teacher in Wellington has just finished scrolling through homes on realestate.co.nz. She's found a two-bedroom apartment in Petone listed at $520,000. She's been saving for three years, has $65,000 in KiwiSaver, and earns $72,000. She has no idea if she can actually afford this.

She Googles "mortgage broker Wellington" and finds your website. She has questions — real, specific questions:

  • Can I use my KiwiSaver for the deposit? How much can I actually withdraw?
  • What's an LVR and do I need 20% or can I do less?
  • I have a $4,000 credit card I haven't paid off — does that matter?
  • How much would my repayments actually be?

She finds a contact form. She fills it in, hits submit, and gets: "Thanks for your enquiry. We'll be in touch during business hours."

She keeps Googling. She finds Squirrel's website, which has a basic chatbot. She tries two other broker websites — both have contact forms only. By 10pm, she's overwhelmed, confused, and has decided to "think about it later." Later turns into never.

This pattern repeats every night across New Zealand. Newpark's data shows that responding within 15 minutes delivers the best conversion rates — 5 to 8 meetings from every 10 leads, converting to 1 to 4 settled deals. After 24 hours, those same 10 leads might produce one meeting.

Around 50% of all NZ home loans are now written through advisers. The market share is strong, but the front door is broken. Most broker websites offer zero conversational engagement outside business hours — exactly when potential clients are browsing.

Why Existing Tools Don't Solve This

Broker CRMs like Trail and MyCRM are excellent pipeline management tools. They track where each client sits in the application process, manage commission calculations, and automate task follow-ups.

Adviser CRM and Tella help manage client relationships and document workflows. Some integrate with lead capture forms on your website.

But none of these tools can have a conversation with a potential client. They capture a name and email. They don't answer the question the person actually has at 9pm on a Thursday night.

Even Squirrel's chatbot — the only conversational tool on a major NZ broker website — handles basic FAQ-level questions. It can't assess a specific person's situation, explain how KiwiSaver First Home Withdrawal works for their deposit amount, or walk them through what LVR tier they'd fall into.

The gap isn't lead capture. It's lead engagement — the moment between "I'm curious" and "I'll book a meeting." That moment happens after hours, and nobody is there for it.

How AI Solves This

A potential client visits your website at 9pm and sees a chat widget. She types:

"Hi, I'm looking at buying my first home. I have about $65k in KiwiSaver and earn $72k. Can I actually afford a place around $520k?"

The AI responds within seconds:

Great question! Let me break this down for you.

KiwiSaver First Home Withdrawal: You can withdraw your voluntary and employer contributions, plus returns — but you must leave a minimum of $1,000 in the account. Based on $65k total, you'd likely be able to withdraw around $63,000-$64,000, depending on your fund's specific breakdown.

Deposit: For a $520,000 property, a 20% deposit would be $104,000. With your KiwiSaver withdrawal plus any cash savings, you might be looking at a low-deposit loan (under 20%). NZ banks can lend up to 95% LVR for first-home buyers, though this depends on the property and your overall financial position.

Rough repayments: At current rates (around 5.5-6%), a $460,000 mortgage over 30 years would be roughly $600-640 per week. But this is very ballpark — the actual amount depends on the rate you secure and your structure.

This is exactly the kind of situation a mortgage adviser can help you map out properly. Would you like to book a free 30-minute consultation with [Broker name]? They specialise in first-home buyers.

The buyer replies:

"Yes please! But I also have a $4,000 credit card — is that going to be a problem?"

Not necessarily — but it does affect your borrowing capacity. Under the CCCFA, lenders look at the credit card's full limit (not just what you owe) when calculating affordability. If the limit is $4,000, they'll factor that into your expenses even if you pay it off monthly.

This is something [Broker name] can advise on specifically — sometimes reducing or closing the limit before applying makes a difference.

I have availability for a consultation this Saturday at 10am or Monday at 6pm (after work). Which works better for you?

The buyer books Monday at 6pm. The AI sends you a WhatsApp summary:

New lead — First Home Buyer Name: Emma Chen | Phone: 021-XXX-XXXX Income: $72k | KiwiSaver: ~$65k | Target: $520k (Petone) Credit card: $4k limit outstanding Consultation booked: Monday 6pm Engagement quality: High — asked specific questions, booked immediately

You arrive Monday morning with a qualified, engaged lead and all the context you need for the meeting.

The Result

  • Every after-hours enquiry gets an instant, knowledgeable response — no more "we'll call you back" dead ends
  • Leads are pre-qualified — you know their income, deposit, KiwiSaver position, and concerns before the first meeting
  • Consultations are booked while interest is high — not 18 hours later when they've moved on
  • You stand out — when every other broker's website has a contact form, yours has a conversation
  • Better conversion rates — responding in seconds instead of hours dramatically increases meeting bookings

What AI Can't Do Here

  • AI won't provide personalised financial advice — it gives general information and always directs to a consultation for specifics
  • AI won't quote exact interest rates or guarantee borrowing amounts — it provides ballpark figures with clear disclaimers
  • AI won't replace the initial consultation — it gets the client to the meeting, where the real advice happens
  • AI can't assess creditworthiness — it captures the information, but the lending assessment is yours

Who This Is For

  • Brokers whose websites have contact forms but no conversational engagement
  • Advisers who specialise in first-home buyers (the demographic most likely to browse after hours)
  • Any broker who checks their enquiries in the morning and wonders how many people visited their site last night and left
  • Broker groups wanting to differentiate their digital presence from competitors

Want This for Your Business?

Book a 45-minute workflow review and we'll show you exactly how this applies to your specific situation — no obligation, no fluff.