Building a mortgage lead follow-up system that survives a busy week
Nobody loses leads on a quiet Tuesday. They lose them in the week three deals are closing, when follow-up is the only thing on the list with no deadline attached to it. A system is what makes that week survivable.
The problem is not effort, it is what gets crowded out
Almost every agent who says they have a lead problem does not have a lead problem. They have a follow-up problem that looks like a lead problem, because the symptom is the same: not enough deals.
The mechanism is boring and consistent. Live files have deadlines — a condition date, a closing, a lender waiting. New leads have none. So in any week where the live files get busy, the leads wait, and by the time you come back to them the enquiry is two weeks old and somebody else has already answered it.
The fix is not discipline, because discipline is exactly the thing that busy weeks consume. The fix is making follow-up something that happens by default and requires a decision to stop, rather than something that requires a decision to start.
Before building anything, count. How many enquiries did you get in the last ninety days, and how many did you contact more than twice? The gap between those two numbers is the whole opportunity, and it is usually larger than anyone expects.
Speed to first contact does most of the work
The first contact attempt matters more than the whole rest of the sequence combined, and the reason is simple: a person who has just filled in a form is thinking about their mortgage right now. An hour later they are back at work. A day later they have filled in two more forms.
You do not need to be instant. You need to be first, and in this industry the bar is lower than it should be.
The practical target is a live human attempt within minutes during business hours, with an automatic acknowledgement covering everything else. The acknowledgement is not the follow-up — it buys you the window in which the follow-up still lands.
- 1
Acknowledge automatically, immediately
The moment a lead arrives, it gets a real acknowledgement that names the person who will call and roughly when. This costs nothing and stops the enquiry feeling ignored while you finish what you are doing.
- 2
Call first, and call fast
Phone beats email badly on a fresh enquiry. Call within minutes if you can, and do it before you have read the whole file — you are trying to reach a person who is currently thinking about you, not to be prepared.
- 3
Route it if you cannot take it
For a team, the lead should reach whoever is actually available, not sit in an inbox belonging to whoever is in a signing. Ownership can be settled afterwards; the first call cannot.
- 4
Log the attempt automatically
If logging a call is a separate task, it will not survive a busy week — and once the log is unreliable, every downstream decision is made on bad data. Calls placed through the system should log themselves.
A cadence that has an ending
Most published follow-up cadences fail for the same reason: they never end, so they are never finished, so they get abandoned. A cadence needs a defined number of attempts, a defined mix of channels, and a defined thing that happens when it runs out.
The shape below is a starting point, not scripture. What matters is that it is written down, that it varies channel, and that it terminates somewhere specific rather than trailing off.
| When | Channel | What it is for |
|---|---|---|
| Immediately | Automated email or text | Acknowledge, name the person calling, set the expectation |
| Within minutes | Call | The attempt that does most of the work |
| Same day | Text | Catches the people who never answer unknown numbers |
| Day 2 | Call at a different hour | Time of day matters more than most people think |
| Day 4 | Email with something useful | A number, a rate, an answer — not a check-in |
| Day 7 | Call | Last live attempt in the active phase |
| Day 10 | Text or email | The polite close-out that invites a reply either way |
| After that | Long-term nurture | Move them, do not drop them — see below |
Every automated message you send has to comply with CASL, and consent has to be real. An enquiry gives you a basis to reply; it is not a licence to add someone to a marketing list indefinitely. Keep the consent basis and the unsubscribe on the record, per contact.
Most of them are not ready, and that is where the value is
The single most valuable habit in this business is treating 'not now' as a date rather than a rejection. Someone renewing in fourteen months is not a dead lead. They are a deal with a known date on it, which is better than most of your pipeline.
This is the part that separates agents who compound from agents who start from zero every January. The book of people who said 'not yet' is an asset that grows every month you work — but only if the not-yets go somewhere with a date attached rather than into a folder.
The mechanism required is unglamorous: capture why not, capture when to come back, and make the come-back automatic. If returning to them depends on you remembering in fourteen months, it will not happen, and you know it will not happen.
- Record the reason: renewing later, house not sold, credit needs work, shopping their bank first.
- Record the date to return, and let the system surface it rather than your memory.
- Keep in light contact in between — rate movement, a market note, something relevant to their reason.
- Re-enter them into an active cadence when the date arrives, automatically.
- Track how many nurtured contacts converted; it is the number that justifies the whole system.
Making it run without you
A written cadence that depends on someone executing it by hand is a wish. The test of whether you have a system is whether it still runs in the week you are heads-down on three closings.
Four things have to be automatic for that to be true. Everything else can be manual.
- 1
Enrolment
A new lead enters the cadence on arrival, by rule, without anybody adding them. The most common leak in every follow-up system is leads that were never enrolled.
- 2
The next action, always visible
At any moment there is a list of who to contact today that assembles itself. Not a report you run — the default screen you open.
- 3
Logging
Calls, texts and emails record themselves against the person. Manual logging is the first thing dropped under pressure, and once it is dropped the pipeline stops reflecting reality.
- 4
The return date
Not-nows come back automatically when their date arrives. This is the piece that turns follow-up from a treadmill into something that accumulates.
If you are testing platforms, ask to see these four specifically, in the product, with a lead you supply. Every CRM will say it does follow-up. Considerably fewer will show you a lead enrolling itself and a not-now returning on its own.
Measure four things, not forty
Follow-up dashboards tend to sprawl into vanity metrics. Four numbers tell you whether the system is working, and each one points at a specific fix when it moves the wrong way.
| Metric | What it tells you | What to do if it is bad |
|---|---|---|
| Time to first contact | Whether speed is holding under load | Automate the acknowledgement and fix routing |
| Contact attempts per lead | Whether the cadence actually completes | Shorten it — an unfinished long cadence is worse than a completed short one |
| Percentage reached at all | Whether your channel mix works | Vary time of day and add a text step before adding volume |
| Nurture-to-deal conversions | Whether the not-nows are coming back | Check the return dates are being set at all — usually they are not |
Count these from the underlying activity records, not from a counter on a campaign. Counters drift, and a drifted number that looks healthy is worse than no number, because you stop looking.
Questions
How many times should I follow up before giving up?
Give up on the active cadence after roughly seven attempts across at least two channels — but do not give up on the person. Move them to long-term nurture with a return date. The active phase should end; the relationship should not.
Is texting mortgage leads allowed in Canada?
Commercial electronic messages are governed by CASL and you need a valid basis for each one, plus a working unsubscribe. An enquiry generally gives you implied consent for a limited window; keep the basis and its expiry recorded per contact rather than assuming it.
Should a team share leads or assign them?
Assign ownership, but route the first attempt to whoever is available. The first call is time-critical and the ownership question is not — settling ownership first is how enquiries end up waiting in somebody's inbox.
Do I need a dialer?
Only if call volume is your constraint. If you are making twenty attempts a day, a phone works. If you are working a list of several hundred, dialling by hand is most of your day, and the logging that follows manual dialling is the part that quietly fails.
What about leads from years ago?
Work them before buying new ones — they cost nothing and you already know something about them. Check your consent basis first, lead with something genuinely useful rather than 'just checking in', and expect the renewal-timed ones to be the best of them.
Bring ninety days of enquiries and we will show you what a completed cadence would have done with them.
See the follow-up system running
Bring ninety days of enquiries and we will show you what a completed cadence would have done with them.