Property leads behave differently from retail orders. One buyer may enquire about several units, a couple may use two phone numbers, and the same lead may arrive from a portal, a Facebook form, a WhatsApp message and a walk-in. Speed matters, but only if the lead reaches the right person with enough context to call.
Automation that works for an online store often fails in this environment. The symptoms below are the ones that appear most often in real estate lead operations, followed by the causes worth checking and the fixes that hold.
Why property leads behave differently
Demand is high-intent but slow to convert. Inventory changes weekly. Agents work from mobile phones and often log activity after the fact. Buyers compare projects, and a lead that looks cold today may be ready next month. Any automation design must preserve context across a long, multi-touch journey rather than optimise for a single transaction.
Symptom: the same buyer is called by three agents
Likely causes: no stable deduplication key; portals resending the same enquiry; separate forms for each project; phone numbers stored with different spacing or country codes; no merge rule when a duplicate arrives.
Fix: normalise phone numbers at intake, create a canonical lead record, define a dedupe window and a merge policy, and lock assignment to one owner until the lead is recycled by a rule. Keep the source history so marketing can still attribute the enquiry.
Symptom: leads are assigned but nobody calls
Likely causes: assignment to a shared queue rather than a person; no response target; the assigned agent is on leave; no escalation when the first call is missed.
Fix: assign to a named agent with a capacity limit, start a response timer at intake, escalate to a team lead if the first contact is not logged, and make reassignment a recorded action. The automation should create a task, not just send a notification.
Symptom: portal leads arrive hours late
Likely causes: polling a mailbox every hour; parsing email formats that change; a webhook that fails silently; no retry when the receiving system is down.
Fix: use a webhook where the portal provides one, put intake behind a queue with retry and a dead-letter path, and monitor lead freshness by source. A simple alert when no lead has arrived from a paid source during business hours will catch more failures than a dashboard nobody opens.
Symptom: WhatsApp follow-up stops or gets blocked
Likely causes: messaging outside the customer service window without an approved template; no recorded opt-in; too many promotional messages; agents using personal numbers.
Fix: capture opt-in at the point of enquiry, use approved templates for initiated messages, keep a human handoff available, cap frequency and centralise the business number. The WhatsApp Business Messaging Policy is explicit about opt-in and template use, and account quality affects everyone who shares the number.
Symptom: site visits and offers vanish between systems
Likely causes: visit notes kept in personal chats; no stage model beyond "lead" and "closed"; no calendar link; no required outcome when a visit is marked complete.
Fix: define stages from enquiry to visit to offer to booking, require a next action and a visit outcome, sync visits to the agent calendar, and report conversion by project and source. A stage that can be skipped without a reason is not a control.
Symptom: reports disagree with the CRM
Likely causes: duplicate records counted twice; different date fields used for the same report; price or area units entered inconsistently; manual spreadsheets layered on top of exports.
Fix: publish one definition per metric, name the source field and the owner, run a nightly reconciliation, and retire spreadsheets that cannot be traced to the CRM. Report on deduplicated leads, not raw rows.
Diagnostic sequence
- Map the journey from first enquiry to booking, including every channel and handoff.
- Count leads at each stage for the last thirty days and compare with the CRM totals.
- Compare intake timestamps with first-contact timestamps by source.
- Sample twenty leads end to end and read the notes as an agent would.
- Check integration logs for failures, retries and duplicate deliveries.
- Interview two agents about the last lead they abandoned and why.
- Fix one break, then re-measure before changing anything else.
Fix patterns that last
- One lead table with a canonical identifier and source history.
- Idempotent intake, so a repeated portal delivery does not create a second lead.
- Named ownership with capacity and an explicit recycle rule.
- Response timers that create tasks and escalate, rather than passive notifications.
- A consent ledger for messaging, with date, source and scope.
- Stage discipline with required fields at visit and offer.
- One reporting model, owned by one person, reconciled nightly.
- A monthly audit of duplicates, stale leads and unlogged activity.
Metrics to watch
- Time from enquiry to first contact, by source and agent.
- Contact rate on new leads.
- Duplicate rate after deduplication rules.
- Visit rate and offer rate by project.
- Share of leads with a next action and a future date.
- Opt-out and block rate on messaging channels.
Where this fits in our work
If your lead flow spans portals, Facebook and WhatsApp, start with Business Automation. The pricing page explains how we scope intake, CRM and reporting work. For the vendor decision that follows, read the AI automation vendor evaluation guide.
Sources
- HubSpot Knowledge Base, Create workflows, accessed 17 November 2026.
- WhatsApp Business Messaging Policy, accessed 17 November 2026.
- DataGuidance, Bangladesh data protection overview, accessed 17 November 2026.
Responses (0)
No comments yet. Add the first one below.
Leave a response
Your email address will not be published. Required fields are marked *.