How a house sale actually runs, task by task, and which parts AI or ordinary automation can take over.
Each task below is marked:
AI: needs a language model, vision model or classifier AUTO: plain workflow rules, scheduling, templates or integrations (no AI needed) HUMAN: keep a person in charge (judgement, trust, legal sign-off, face-to-face) Stage 1: Win the instruction (before listing)
Task
Type
How to automate
Enquiry capture from website, portals and phone AUTO + AI Web forms and portal leads push straight into the CRM . An AI chatbot or voice receptionist answers out of hours and books valuations. Booking the valuation / listing appointment AUTO Online booking calendar with confirmation emails and reminders. Gathering property facts (tenure, lease, council tax band, EPC, flood risk, planning, title) AUTO + AI Pull open data (Land Registry, EPC register, flood maps) by postcode. AI drafts a fact sheet from the seller’s answers. Comparable sales analysis (CMA) AI Model compares recent sold prices, adjusts for size, condition and features, and drafts a price range for the agent to check. Local market research (days on market, price trends) AUTO Scheduled data pulls into a dashboard. Listing presentation deck AI + AUTO Templates filled from CRM data; AI writes the narrative on the market. Instruction paperwork (sole agency / terms, property information questionnaire) AUTO E-signature with pre-filled fields from the CRM. Anti-money-laundering and ID checks AUTO Digital ID verification service, results logged to the file. Follow-up after the appointment if no decision AUTO + AI Timed email sequence; AI personalises it using notes from the appointment.
Stage 2: Prepare and market the property
Task
Type
How to automate
Booking photographer, floorplan and video AUTO Booking link with the photographer’s calendar; job auto-created on instruction. Culling and editing photos AI Auto-select the best shots, straighten and brighten them. Virtual staging and decluttering suggestions AI Generate staged images of empty rooms; vision model gives a room-by-room list of improvements. Writing the listing description and headline AI Draft from the fact sheet and photos, then agent edits. Check for banned phrases and fair-housing / property misdescription rules. Publishing to Rightmove, Zoopla, OnTheMarket and the agency website AUTO Single listing record pushed to portals via feed or syndication tool. Price changes and status updates across all portals AUTO One edit in the CRM updates every channel. Social media posts and short video AI + AUTO AI makes post variants and captions; scheduler publishes. Email campaign to buyers on the database (“just listed” ) AUTO Segmented list by criteria (budget, area, bedrooms), templated HTML email. Brochures and flyers AUTO Generated from the same listing data. Board and lockbox setup HUMAN Physical job; log the install in the CRM. Vendor marketing report AUTO + AI Weekly report built from portal views, enquiries and viewings, with an AI-written summary.
Stage 3: Viewings and enquiries
Task
Type
How to automate
Answering buyer enquiries AI Chatbot or email assistant answers facts (price, size, availability, parking) from the listing data; hands complex questions to a person. Qualifying buyers (budget, chain, mortgage in principle, timing) AI + AUTO Short form or chat flow; score the lead; flag to the negotiator. Booking viewings around the vendor’s availability AUTO Self-service booking with rules (buffer times, no-go slots, key holder availability). Sending viewing confirmations and directions AUTO Templated emails and SMS with reminders. Accompanied viewings HUMAN Keep a person here. Post-viewing feedback from buyers AUTO + AI Automated request after each viewing; AI summarises the feedback into a themed report for the vendor. Vendor feedback reports AUTO + AI Combines feedback, sentiment and portal data; flags patterns (e.g., “kitchen mentioned in 4 of 6 viewings” ). Price reduction advice HUMAN + AI AI flags listings with falling interest; agent makes the call.
Stage 4: Offers and negotiation
Task
Type
How to automate
Logging offers received AUTO Offers entered by email parsing or form; time-stamped in the CRM. Net sheet and offer comparison AUTO + AI Calculates net proceeds per offer; AI checks chain strength and funding risk. Negotiation strategy and vendor advice HUMAN Keep a person here. Drafting offer acceptance and conditions AI + AUTO Generated from templates; agent reviews. Memorandum of sale / sale agreed notifications AUTO Triggered on acceptance: emails to all parties, portal status changed to “Sold STC” .
Stage 5: Sale agreed to completion (sales progression)
Task
Type
How to automate
Referring to conveyancer and mortgage broker AUTO Referral forms with the file details already filled in. Chasing searches, mortgage offer, ID and deposit AUTO Automated reminders to each party, with a shared status tracker. Reading and summarising legal documents AI Extract key facts from title deeds, leases and management packs; flag unusual clauses for the conveyancer. Chain management (linked sales) AUTO Chain tracker showing every link’s status; alerts when one breaks. Keeping buyer and seller updated AUTO + AI Scheduled progress updates with an AI summary of what is outstanding. Key handover and completion day checklist AUTO Checklist and meter reading reminders. Final invoice and commission AUTO Invoice generated on completion from the sale record.
Stage 6: Admin, compliance and growth
Task
Type
How to automate
Record keeping and file audit AUTO Automatic logging of every email, call and document to the property file. Data protection (GDPR) requests and consent AUTO Consent capture on forms; deletion requests routed to one owner. Compliance checks (Property Ombudsman, client money, complaints log) AUTO Checklist per file; alerts for missing items. Reviews and referrals AUTO Review request sent after completion. Reporting (pipeline, conversion, time-to-sell) AUTO + AI Live dashboard; AI writes the weekly summary.
Summary: what to automate first
Highest value, lowest risk:
Enquiry capture, booking and reminders (AUTO) Listing syndication and price-change updates (AUTO) Listing description and social post drafts (AI, agent edits) Chasing and progress updates in sales progression (AUTO) Post-viewing feedback collection and vendor reports (AUTO + AI) Document summaries for the conveyancer (AI, human checks) Keep a person on:
Valuation judgement and pricing advice Negotiation Accompanied viewings Legal advice and anything that needs a signature decision Client relationships at the key moments Caveats
Automation does not remove the agent’s legal duties (client money, AML checks, property misdescription rules). Build those checks in and keep a record. AI valuations and summaries must be checked by a person before they reach a client. Time savings depend on the agency’s volume and current tools. No figures are claimed here. Sources consulted
NAR, “179 Ways Agents Who Are REALTORS Are Worth Every Penny of Their Compensation” (nar.realtor) McKissock, “Master the Process: How to Sell a House as a Real Estate Agent ” Pipedrive, “Real Estate Sales Process: Steps and Guidelines” Opendoor, “How to Sell Your House in 2026” Coldwell Banker Heritage, “The Home Selling Process, Step by Step”
