The problem nobody talks about honestly
Look at your Google Analytics. Filter to property-detail page views. Split by hour-of-day. You will see two peaks — one between 8 and 9pm on weekdays, one across Saturday and Sunday afternoons.
That's your real audience. Working professionals, school-run families, dog walkers — all browsing your stock when they actually have time, which is exactly when your office is shut.
In our data, 73% of UK property searches happen outside 9-5. That number has been climbing for five years and isn't going back down.
What happens to those visitors? Three possibilities:
- They use your contact form and you reply on Monday morning, by which time they've messaged six other agents.
- They get your voicemail and don't leave a message.
- They close the tab.
All three are losses. The lead-to-conversion data says it pretty clearly: the agent who replies first wins between 40-60% of the time. Monday morning is too late.
What "after-hours capture" actually means
You don't need a 24-hour call centre. You don't need to staff nights. You don't even need to appear to be open. What you need is a way for a visitor at 11pm to:
- Get the information they came for (property details, area info, valuation estimate)
- Leave their contact details so you can ring them in the morning
- Feel that they've been heard, not that they've been ghosted
That's a three-job system. Let's look at each.
Job 1: Answer the question they came for
A buyer browsing your 3-bed semi in Wimbledon at 10pm wants to know two things: is it still available, and what's the area like.
A static property page gives them the listing but not the answer to "is it still available?" (price reductions, sold subject to contract, withdrawn — your CRM knows, your website doesn't). An AI assistant connected to your live feed can answer that in seconds.
For area questions, the answer is the same: pull from real data sources (schools, crime, broadband, flood risk) and answer immediately. Don't make the visitor click around the council's website.
Job 2: Capture them before they bounce
The single highest-converting moment in a property browsing session is right after the visitor's question has been answered. They're engaged, they've invested in the chat, and they want more.
This is the moment to ask for the email — not as a form gate, but as a natural next beat:
"I've got the figure ready. What's the best email for our records?"
That phrasing matters. "For our records" signals professionalism. "I've got the figure ready" frames the ask as part of delivering value, not extracting data.
Same goes for phone numbers. Pair the ask with two reasons:
"And a mobile so the team can confirm the figures or arrange a proper visit."
Two reasons in one breath. The visitor picks whichever resonates and shares the number anyway.
Job 3: Score the lead so your morning is efficient
Not every after-hours capture is hot. Some are vendors fishing for a number, some are tenants looking to rent in 18 months, some are competitors. Your team's time on Monday morning is your scarcest resource — burning it on cold leads is worse than not having them.
A decent lead-scoring layer looks at:
- Engagement depth. Did they only ask one question, or did they go through five?
- Specificity of intent. "Looking to buy" is weak; "want to view 22 Acacia Avenue this Saturday" is strong.
- Information depth. Did they share both email AND phone, or just an email?
- Time spent. A 30-second exit is a different animal from a 6-minute conversation.
Score them, sort them, ring the hot ones first.
The economics
Let's do the maths on a typical UK independent.
- 800 monthly website visitors
- 40 of those generate an enquiry currently (5%)
- Of those 40, you convert 4-6 into instructions (10-15%)
Adding 24/7 enquiry capture typically lifts the enquiry rate to 12-18% (because you're catching the 73% you currently lose). For 800 visitors, that's 100-140 enquiries instead of 40 — call it 3x.
Even if your conversion rate to instruction drops a bit because the new leads are colder (call it 8-10%), that's 8-14 instructions per month instead of 4-6. At an average UK sales fee of £4,500, that's £18-36k of incremental fee income per month from work that didn't require staffing nights.
The maths usually beats the £59-129/month subscription by an absurd margin. Even on a bad month.
What to look for in a setup
Three things to test before you commit:
- Real listings only. The bot must quote your actual stock by reference, never invent. A bot that hallucinates "we have a great 2-bed in Wimbledon for £400k" is a lawsuit waiting.
- Source-linked answers. If a visitor asks about flood risk and the bot says "the area is low risk", it should be answering from a real data source, not the model's training data. You need to be able to defend the answer.
- Lead scoring built in. Without it, you're back to triaging manually on Monday morning. Which defeats the point.
The five-minute setup
A good system shouldn't cost you a working day to install. The norm should be:
- Paste your domain URL → website scraped, basic Q&A ready in minutes
- Connect your CRM feed (when you're ready, not gating launch)
- Drop a script tag on your homepage
- Set your brand colour + logo
That's it. If a vendor is selling you a six-week implementation project, ask why. Most often it's because their product needs hand-holding to actually work.
The bottom line
The after-hours hours of your website are doing more work than your office is. Catching that traffic doesn't require staffing nights, doesn't require a CRM rebuild, and doesn't require a six-week pilot. It requires a single, well-built chat assistant that does three specific jobs.
The investment is minimal. The return is the difference between Monday voicemails and Monday hot leads.