Selling an apartment consists of two very different halves. The first is answering the questions everyone asks — price, handover date, amenities. The second is persuading, addressing hesitation and closing. A robot does the first half well and the second half not at all, and knowing that boundary is the whole story.
How much of it repeats
Log the questions visitors ask in a sales gallery for one week. The result is usually surprising.
A very large share are the same: unit sizes, indicative pricing, construction progress, payment terms, location, amenities, parking, financing arrangements.
Sales staff answer these dozens of times a day. Answer quality declines through the day, and the time spent leaves less for visitors with genuine intent.
This is the part a robot absorbs completely. Fixed content, no judgement required, answered identically on the fiftieth occasion as on the first.
An overlooked second benefit. Beyond saving time, a robot ensures every visitor receives identical information. In property sales, staff giving different answers about the same detail is a genuine source of complaint.
So the first value here is not novelty. It is consistency of information.
Dividing roles with sales staff
This determines whether the robot is welcomed or treated as a threat.
The robot takes the front end. Greeting, general project introduction, basic questions, walking visitors through the scale model, and most importantly establishing which category the visitor falls into.
Staff take the back end. Needs-based advice, addressing hesitation, negotiation, closing.
A clear transfer point. When a visitor asks specific financing questions, shows serious interest, or asks about a particular unit — transfer immediately.
Information must follow the visitor. The agent receives what has been asked and which unit type interests them. Without this, the robot has only cost the visitor time.
Be explicit with the sales team. The robot does not take their customers; it filters out browsers so they can focus on serious buyers. If the sales team does not understand this, they will stand in front of the robot.
Content for property robots
This sector has content requirements others do not.
Information must be exactly right. Areas, prices, timelines, legal status. A wrong figure is a real risk because visitors remember and may record.
Distinguish confirmed from indicative. Expected handover, reference pricing, planned amenities. The robot must state the level of certainty rather than assert.
No commitments. The robot must never commit on price, discount or timeline. Configure it to transfer on every question of that kind.
Update with sales progress. Which units remain, which policies currently apply — these change weekly, and stale information causes real problems.
Legal content should be approved. Anything about a project's legal status should pass the legal team before loading.
The safety principle: the robot provides information, people make commitments. This boundary must be configured firmly rather than relied on the robot to observe by itself.
Capturing visitor information
The most measurable value and the part needing careful design.
Give visitors a reason to leave details. A detailed price list, floor plans, notification of the next release. Asking for information without offering anything produces very low consent rates.
Ask for little. A name and one contact channel suffices. Long forms are abandoned midway.
Record what interested them. Unit type, price range, expected purchase timing. This is more valuable than a phone number because it shapes how the lead is followed up.
Push it into the sales system immediately. Information sitting inside the robot that nobody extracts does not exist.
Respect privacy. State clearly what the information is used for and obtain consent. This is both a legal requirement and something customers increasingly care about.
Measure the conversion chain. Visitors who spoke to the robot, those who left details, those contacted by an agent. That sequence shows exactly where the drop-off is.
Measuring in a sales context
Property has long sales cycles, so interim metrics matter rather than waiting for transactions.
Visitor interactions. The basic figure. If low, the issue is where the robot sits in the gallery.
Contact details captured. And the ratio against total visitors.
Staff time saved. Estimated as basic questions handled multiplied by average time each.
Time visitors spend in the gallery. A robot holding visitors longer creates more opportunities for staff contact.
Sales team feedback. The single most important indicator. If they find it helpful it will be used; if they find it obstructive it will be neutralised in various ways.
Photographs and shares. For a project still building recognition, this communications value may exceed the direct sales value.
Frequently asked questions
What is the first value of a robot in a sales gallery?
Consistency of information. Beyond saving staff time, it ensures every visitor receives identical answers — and staff giving different answers about the same detail is a genuine source of complaint in property sales.
May a robot commit on price?
Never. The safety principle is that the robot provides information and people make commitments, and that boundary must be configured firmly rather than relying on the robot to observe it by itself.
How do you keep the sales team from seeing it as a threat?
Explain that it filters out browsers so they can focus on serious buyers, and ensure the information a visitor already gave transfers to the agent so nothing is repeated.
What should be measured with long sales cycles?
Interim metrics: interactions, contact details captured, staff time saved, visitor dwell time, and above all the feedback of the sales team itself.
More in Training and customising a robot and Deployment and operations.
