How one Essex agency cut enquiry response from 6 hours to 2 minutes
A lettings and sales agency in Essex was getting around 140 enquiries a week across three portals and their own site, split across four branches. One negotiator was covering first response for the entire patch. Their own numbers showed the average first reply took 6 hours during the week and over 20 hours if an enquiry came in on a Saturday afternoon, because nobody checked the inbox again until Monday.
After setting up AI automation for first response, viewing booking, and basic FAQ handling, average first response dropped to under two minutes. Viewings booked without a human touching them went from zero to roughly 35% of all bookings. The negotiator who used to spend her mornings answering “is this still available” messages was freed up to call warm leads asking about mortgage in principle or move-in dates.
That is the outcome. Here is how it actually works.
What the Automation Does, Step by Step
The flow is the same whether an agency builds this themselves or brings in outside help. An enquiry lands from Rightmove, Zoopla, OnTheMarket, the agency’s own website, or WhatsApp. The system pulls listing details from the CRM (Reapit, Jupix, Alto, and Street are the common UK platforms) so the reply is specific, not a generic acknowledgement.
Within one to two minutes, the enquirer gets a reply that names the property, the street, the council tax band, the pet policy, and three real viewing slots pulled from the negotiator’s actual calendar. Not a made-up availability list. If they reply to confirm a slot, it books straight into the calendar without anyone touching a keyboard.
- Enquiry arrives from portal, website form, or WhatsApp
- AI pulls live listing data from CRM
- Personalised first reply goes out within one to two minutes
- Real viewing slots from the negotiator’s calendar are offered
- Booking confirmation sent by text and email, with a 24-hour reminder
- If no reply within a day, follow-ups go out on day one, day three, and day seven
- Lead is marked cold and removed from the human’s priority list after the sequence
- Anything that looks like a serious buyer or a complaint gets flagged to a person immediately
The Number That Actually Surprised the Branch Manager
The speed improvement was expected. What caught the branch manager off guard was how many enquiries turned out to be dead the moment someone replied. People who had enquired on five flats simultaneously and had already found somewhere else. Automation did not create those tyre kickers. It found them faster, so the negotiator stopped chasing them for a week.
That is the real efficiency gain. Not manufacturing new buyers. Stopping the leak of buyers you already paid Rightmove and Zoopla portal fees to attract.
One well-known lead response study found that contacting a lead within five minutes makes you roughly 21 times more likely to qualify them than waiting 30 minutes. Agents read that and assume AI will win them more deals. What it actually does is stop losing deals that were already half won.
What It Costs in the UK
Rough UK pricing as of 2026: a basic AI enquiry handling setup built on an existing CRM integration runs between £150 and £400 a month for a single branch, depending on enquiry volume and portal coverage. A multi-branch agency building something more custom, with WhatsApp, voice call handling, and CRM write-back, is looking at an upfront build cost of £2,000 to £6,000 depending on complexity, plus a smaller monthly fee for hosting and support.
The comparison point is portal fees. Rightmove and Zoopla combined often run £800 to £1,500 a month per branch. If automation recovers even one extra instructed property a month that would otherwise have gone to the agent who replied faster, the maths works. That is the frame worth using, not the demo video.
A single busy branch handling 30 to 50 enquiries a week can still see a return, mainly through weekend and evening cover. Most missed enquiries happen outside office hours when nobody is checking a portal inbox at all.
⚠️ Where the Automation Has to Stop
This is where agencies get it wrong most often, by trying to make the AI do too much. There are clear lines it should not cross:
- Never negotiate price. That is a human job.
- Never handle a complaint about a damp survey or a broken boiler. An AI arguing with a tenant at 11pm will lose the landlord.
- Never pretend to be a named negotiator who does not exist. If someone directly asks whether they are talking to a bot, the system should say so. Losing that trust with someone about to sign a 12-month tenancy is a bad trade.
- Never run unsupervised on stale data. Property status changes (sold subject to contract, landlord pulls out) and an AI replying with outdated availability makes the whole agency look sloppy.
The cleanest way to think about the boundary: automate the repeatable, factual, high-volume first step. Keep a human on anything that involves judgement, money, or an upset person. Property enquiries are almost entirely the first category. Complaints and negotiations are almost entirely the second.
The One Practical Takeaway
The agencies doing this well are not the ones with the flashiest chatbot. They are the ones who worked out exactly which three questions get asked a hundred times a week, automated those, and left everything else to the people who are good at their jobs.
A system that answers “is this still available,” offers a viewing slot, and books it into a calendar without human input is not replacing a negotiator who knows the street and can read a buyer’s tone on the phone. It is removing the hour of inbox triage that used to happen before that negotiator could start doing their actual job.
Frequently Asked Questions
Do sellers and landlords know when AI is handling the first reply?
Most agencies do not advertise it but do not hide it either. Good practice is to disclose it if directly asked and to make sure the AI speaks as “the team at [agency name]” rather than as a named negotiator who does not exist.
Can it double-book a viewing slot?
Not if it is set up correctly. A properly integrated system reads real availability from the negotiator’s calendar or the branch booking system before offering slots, so double bookings do not happen.
What happens when a property has already gone under offer?
A well-built system pulls live status from the CRM before replying and tells the enquirer the property is under offer, then offers similar listings instead. This requires the CRM and the AI tool to be connected. Many cheap off-the-shelf chatbots fall down here.
Is this only worth it for large agencies?
No. A single branch handling 30 to 50 enquiries a week can see a real return, primarily through coverage outside office hours when most enquiries go unanswered.
I’m founder of BizStack at Cagri Sarigoz LLC and a passionate advocate for entrepreneurs.
With over 14 years in tech, marketing, and AI, including my role as Head of SEO at CitizenShipper and co-founder of TaleBot at Intale AI, I’m dedicated to sharing genuine, useful product insights and tips.
At BizStack, I aim to cut through the digital noise to provide clear, actionable advice.
And more than all else, I’m a father to a (always) little girl and a husband.
Contact me at [email protected] for assistance.
