AI lead qualification for Zoopla and OnTheMarket enquiries

Agencies receiving Zoopla and OnTheMarket enquiries face a familiar pattern: the buyer sends a message, the inbox is busy, and the reply goes back hours later. By that point, the buyer has spoken to someone else. Zoopla OnTheMarket AI lead qualification addresses this window directly. The enquiry arrives, the agent reads the property record, and a qualifying reply goes back within seconds. Not a template. A specific response, in your agency's voice, to what the buyer actually asked.

This post explains how the agent works, what it changes on a busy Monday morning, and how to build it in a way that keeps you on the right side of UK GDPR and the TPO code.

Why Zoopla and OnTheMarket enquiries are hard to handle well

Zoopla and OnTheMarket each generate their own enquiry flows. Both land in the agency inbox by email. Both carry the same problem: there is no relationship behind the enquiry yet. The buyer has seen a photograph, read a short description, and decided to make contact. That contact costs them almost nothing, which means the quality of the enquiry pool varies widely. Some are proceedable buyers ready to view this week. Others are browsing with no timeline. A few are already under offer on something else and are keeping options open.

The job of qualification is to separate these groups quickly, without losing a good buyer to a slow reply. Most agencies handle it the same way: a negotiator reads the enquiry when they reach it, decides whether it looks like a real buyer, and either calls or sends a holding response. The holding response is often a template. It does not qualify the buyer, and it does not advance the conversation.

The result is a funnel where too many good enquiries sit in limbo. Portal behaviour punishes this. Buyers who send enquiries to multiple properties, or to multiple agencies, will speak to whoever replies first. If that first conversation is substantive and helpful, the buyer tends to stay with that agent. If it is a template sent three hours later, the buyer moves on.

Zoopla and OnTheMarket both allow agencies to track enquiry volumes and response patterns. Most agencies find, when they look at the data, that their average first-reply time is considerably longer than they expected.

How AI lead qualification works on Zoopla and OnTheMarket enquiries

The architecture is an event-triggered agent with one clear loop: trigger, read context, reason, take action, log.

The trigger is an inbound enquiry from Zoopla or OnTheMarket. The agent connects to your enquiry flow via webhook or email parsing, depending on what the portal makes available. It receives the enquiry the moment it arrives, before the inbox is opened.

The agent reads the enquiry text, then retrieves the property record from the CRM. It checks the property type, price, tenure, EPC rating, current availability, viewing history and any notes already on file. It also reads any earlier exchange with the same buyer, if one exists, so the reply is continuous rather than starting cold.

With that context assembled, the agent decides on the right next step. A buyer who has asked a specific question about the garden size gets a direct answer. A buyer who has sent a short "I'd like more information" message gets a reply that includes the most relevant detail from the property record and asks two or three qualifying questions: buying position, current chain status, preferred viewing timescale. A buyer who has already provided those signals and looks proceedable gets a viewing invitation with a choice of slots pulled from the live diary.

Every reply goes back in your agency's voice, not a generic AI tone. The voice, vocabulary and sign-off are agreed with your team before the agent goes live and tested against a sample of real enquiries from your inbox before it speaks to any live buyer.

If the buyer replies and the conversation develops, the agent continues it. Booking confirmed, diary updated, CRM note written. If the buyer says something the agent cannot handle from available data, or raises an objection that needs a negotiator, the thread is escalated. The negotiator receives a summary of the conversation to date alongside the escalation flag, so they pick up a briefed exchange rather than a cold thread.

The same loop that handles Zoopla and OnTheMarket enquiries also powers Sortd's lead qualification service. Once a lead converts to a booked viewing, the viewing receptionist takes over: handling confirmation, rescheduling and viewing-day reminders without negotiator input. And the buyer context gathered during qualification, buying position, timeline, chain status, feeds directly into sales progression if an offer follows, so the progression agent starts with that context already assembled.

What it looks like for an agency

Consider an agency with one branch running Zoopla and OnTheMarket alongside Rightmove. On a Friday afternoon, four enquiries arrive between 5pm and 8pm. Previously, all four would wait until Monday morning.

With an AI lead qualification agent running, the pattern changes.

The first enquiry arrives at 5.15pm. It is a two-line message asking whether the property has a garage. The agent checks the property record, finds the relevant detail, and replies with a clear answer, confirmation that the property is still available, and two qualifying questions. The buyer replies at 7pm: mortgage-ready, end of chain, wants to view Tuesday. The agent books Tuesday at 2pm, sends the buyer a confirmation and logs the exchange to the CRM.

The second enquiry is vaguer. The buyer is clearly early-stage and has not answered any qualifying questions. The agent replies with property detail and asks a clear question about timeline. No further response arrives before Monday. The agent flags it as an unresolved thread.

The third and fourth enquiries qualify at different speeds through the weekend. By Monday morning, two viewings are booked, one thread is flagged and waiting, and one has been escalated because the buyer mentioned difficulty getting a decision in principle.

The negotiator opens the CRM on Monday to find a clear picture rather than a backlog of unanswered messages.

UK GDPR, TPO and integration in practice

Portal enquiries carry personal data from the first contact. Name, email address, phone number and anything the buyer has chosen to share about their circumstances. UK GDPR applies from that first message. The lawful basis for processing is typically legitimate interests: the buyer chose to make contact and a response is required to fulfil that interaction. The agent processes only what is necessary for that purpose and does not retain data beyond what the enquiry exchange requires.

Sortd builds use UK or EU data residency by default. No buyer data routes through US-based infrastructure. Each build includes a Data Processing Agreement, and the agency's existing privacy notice is reviewed to confirm that AI-assisted responses are covered. The ICO's guidance on UK GDPR sets out the controller and processor obligations clearly.

The TPO code requires that every communication in the agency's name is accurate and retrievable. Every AI-generated reply is logged, time-stamped and linked to the enquiry record. The agency can pull any message in the event of a complaint. The record is more complete than an email sent manually without a corresponding CRM note.

On integration, the agent connects to the main UK CRMs: Reapit, Alto, Jupix, Vebra, Dezrez and agentOS. Zoopla and OnTheMarket enquiries arrive via webhook or email hook depending on each portal's available setup. The Negotiator covers how portal technology and AI adoption are developing across the UK agency market if you want a broader view.

How to get started

The starting point is a review of your current portal enquiry handling. How many enquiries arrive each week across Zoopla and OnTheMarket? What is your average first-reply time? What proportion convert to a booked viewing, and how many go cold before any meaningful exchange takes place?

Those numbers establish the baseline. A well-built lead qualification agent should move all three in a measurable direction. If it does not, it is not configured correctly.

Sortd's process is a discovery call followed by a free working version of the agent, built against your own CRM data and tested on a sample of real enquiries. You see how it replies to your actual leads, in your agency's voice, before it speaks to any live buyer.

If Zoopla and OnTheMarket enquiries are going cold before your team reaches them, start with a conversation.