AI prospecting
How AI propensity-to-sell models are changing estate agent prospecting
Instead of blanket-mailing an entire postcode and hoping for a response, propensity-to-sell modelling helps estate agents identify the homeowners statistically more likely to list — and focus their prospecting budget accordingly.
At a glance
Predict who may move: propensity models analyse property data, sales history and local market conditions.
Cut wasted prospecting: spend is concentrated on households more likely to instruct.
Keep local expertise: agents choose their target market first, then layer AI predictions on top.
Connect the journey: propensity data can feed into direct mail, CRM and nurture activity.
Prospecting has always been a numbers game for estate agents.
Print enough letters, knock on enough doors and some percentage of homeowners will respond.
The problem is that the percentage has historically been small, while the cost of reaching everyone who is not interested continues to add up. Every letter sent to a homeowner with no intention of moving is budget spent with little chance of a return.
Propensity-to-sell modelling changes that equation by identifying which homeowners are statistically more likely to list in the near future — before they have contacted an agent.
Instead of asking “which postcode should we mail?”, the question becomes “which homes inside that postcode are most worth targeting?”
What data goes into a propensity-to-sell model?
Spectre says its AI draws on more than 1 billion data points across the UK’s 29 million residential properties.
The model combines property-level information with historical sales data and market trends to produce a prediction for each property, taking account of current local market conditions as well as longer-term patterns.
Ownership length is one of the strongest signals. A homeowner who bought eight years ago may be statistically more likely to sell than somebody who moved 18 months ago — but the model goes considerably further than that single factor.
Signals can include:
- How long the current owner has held the property
- Historical sales patterns
- Hyper-local listing activity
- Changes in local property values
- Whether nearby homes have recently come to market
- Whether a property has previously been listed and withdrawn
Spectre says the predictions are continually trained on current market data and validated against what subsequently happens, so the model can adjust to actual market behaviour rather than relying on static assumptions.
How does predictive targeting change prospecting economics?
Traditional estate agency prospecting generally follows a broadcast model.
An agency chooses an area, prints thousands of letters and sends them to every residential address within its target zone. The majority of recipients may have no intention of moving.
Predictive targeting turns that approach around. The agent starts with the households most likely to come to market and concentrates spend on reaching those homes.
Lower volume. Better targeting.
The goal is not simply to send more prospecting. It is to spend more of the budget reaching households with a stronger statistical likelihood of moving.
That can affect more than printing and postage. It also concentrates negotiator follow-up time on prospects the data suggests may be closer to making a move.
According to Spectre’s own testing, adding its propensity predictions to campaign targeting criteria increases ROI by an average of 310%.
Spectre testing
+310%
Average increase in campaign ROI reported by Spectre when its AI predictions are added to targeting criteria.
How does Spectre Sales use propensity data in campaigns?
The Spectre Sales campaign builder gives agents a range of filters for shaping a direct-mail audience around their own local knowledge and commercial priorities.
Agents can target using criteria including estimated property value, time since last sale, property type, location and other property-level information.
The process happens in two stages:
1. The agent builds the audience.
The agency uses its own market knowledge and Spectre’s campaign filters to decide which properties it wants to prospect.
2. AI sharpens the list.
Spectre AI can then remove properties with a lower predicted likelihood of instructing and retain those with the strongest propensity scores.
The agent remains in control of which properties to target. The AI predictions act as recommendations designed to sharpen an already targeted campaign rather than replace local expertise.
Local knowledge decides where to play. AI helps decide which doors may be most worth knocking on.
Automating the prospecting agents already know works
Spectre Sales also automates 20:20 campaigns around every property an agent has sold, helping agencies maintain a consistent presence around previous instructions.
Alongside this, the platform includes what Spectre describes as “anti-embarrassment” safeguards designed to prevent badly timed prospecting.
Campaign exclusions can automatically remove:
- Current agency clients
- Recently completed transactions
- Properties already on the market with the agency
That helps avoid the awkward prospecting letter arriving at the home of a vendor whose property the agency is already selling, or a previous client being cold-contacted only weeks after completion.
Do propensity models actually work in practice?
Scepticism around predictive modelling in property is understandable. Estate agency remains a relationship business and no algorithm can say with certainty that one particular homeowner will sell within a particular timeframe.
The point of propensity scoring is different. It identifies statistical patterns across a large number of properties, allowing an agency to prioritise households that display more of the signals associated with future instructions.
National agency case study reported by Spectre
124,527
targeted letters sent
167
instructions generated
£345m
combined listing value
23x
reported return on campaign spend
The model is not claiming certainty at an individual-address level. The commercial advantage comes from applying statistical likelihood across hundreds or thousands of prospects, where relatively small improvements in targeting can translate into meaningful gains in campaign performance.
How does AI prospecting fit into a wider marketing strategy?
Propensity-to-sell modelling becomes more useful when it is connected to the rest of the agency’s marketing rather than treated as a standalone prospecting tool.
When Spectre Sales sits alongside Street.co.uk CRM and the wider Spectre marketing suite, the tools can use a shared data layer and coordinate follow-up communications across channels.
For example:
A homeowner receives a targeted prospecting letter.
They later submit an instant valuation request through the agency website.
The lead flows into Street CRM.
A Spectre Email nurture journey can then be triggered automatically to continue the conversation.
Likewise, a homeowner who does not respond to the first letter can potentially be re-targeted later if their propensity score increases as local market conditions change.
The result is a more connected approach where direct mail, CRM activity and nurture communications are informed by the same underlying data.
From broadcast prospecting to data-informed timing.
AI does not remove the need for good creative, strong follow-up or local knowledge. It helps agencies decide where those resources may have the greatest chance of producing an instruction.
A different way to think about prospecting
The shift from broadcast to targeted prospecting is already changing the way agencies think about direct mail.
Rather than measuring success by how many doors an agency can reach, propensity-to-sell models encourage teams to think about how accurately they can identify the households worth prioritising.
For agents, that means the opportunity is not simply to prospect more. It is to combine local expertise with better data, reduce wasted activity and focus marketing spend on the homeowners statistically more likely to make a move.
Spectre Sales
Put your prospecting budget behind the properties that matter most
See how Spectre Sales combines granular campaign targeting with AI propensity predictions to help estate agents refine prospecting and reduce wasted spend.
