Estate agents could be overlooking potential instructions and revenue already sitting within their CRM systems. New research suggests many agency leaders believe their historic databases are underused.
A survey conducted by Kerfuffle found that 82.4% of agency leaders believe their existing databases contain untapped revenue.
More than a quarter – 26.5% – said they know potential revenue exists within their data but struggle to find it.
The findings highlight a longstanding challenge for estate agencies. Branches can accumulate thousands of records covering vendors, landlords, buyers, tenants, valuations and previous transactions.
However, as databases grow, agents can find older information difficult to search and use effectively. Records can also become outdated, duplicated or poorly structured.
That can leave agents spending money generating new leads while existing contacts and historic valuation opportunities remain untouched.
CRM migration remains a concern
The issue becomes particularly important when agencies change CRM providers.
According to the Kerfuffle research, 58.8% of agency bosses fear losing or burying vital historical information when migrating systems.
For established agencies, the volume of information involved can be substantial.
Greenhouse OS, which provides CRM technology to estate agencies, recently migrated more than 1.2 million records for a Hertfordshire agency. The seven-branch business has been trading for more than 45 years.
A separate migration for a three-branch Essex agency involved 748,831 records accumulated over 20 years.
The figures demonstrate the scale of data that established agencies can hold. They also highlight the challenge involved in transferring that information between systems.
AI increases the value of historic agency data
The growing adoption of artificial intelligence across estate agency is putting greater emphasis on CRM data.
AI systems can analyse historic property and customer information to identify patterns and potential opportunities. However, their effectiveness depends heavily on the quality of the underlying information.
Ian Preston, CEO of Greenhouse OS, believes one potential application is using historic data to personalise customer communications at scale.
He said: “The biggest single application of AI in estate agency right now is hyper-personalisation, making something delivered on a mass scale feel like it was created just for one person.
“But to power that, you need rich history.”
Historic records can include previous valuations, transactions, landlord relationships and buyer enquiries. That information can help agents identify previous contacts who may be approaching another property decision.
Technology can also automate parts of that process. CRM and AI systems can identify and prioritise old contacts for follow-up.
For agents, the question is no longer simply how much information their CRM holds. Increasingly, it is whether the business can make effective use of that data.
Comments (3)
The size of the database is only part of the question. Before an AI system decides who deserves a call, the record needs enough context to explain why: when the information was last verified, what the person previously wanted, whether that reason still applies, and whether the agency can properly use the data for this contact.
That makes the useful pilot measures more demanding than messages sent or records scored. I would want to see the proportion of suggested contacts that became qualified conversations, the stale or duplicated record rate, opt-outs or complaints, and how often a negotiator rejected the suggested reason for contact. Those measures reveal whether the system is recovering genuine opportunity or simply making old data easier to message at scale.
Migration matters for the same reason. A flat note saying ‘historic valuation’ is much less useful than a dated sequence of events with source, outcome and ownership preserved. AI can rank the next action, but only if the underlying history survives the move in a form that can be checked.
This is exactly the challenge, and I completely agree that simply making old data easier to message isn’t the objective.
For me, the key distinction is between scoring a record and understanding an opportunity.
A 2021 valuation enquiry shouldn’t rank highly simply because “valuation” appears in the record.
PropFlow needs to understand the sequence: what happened, when it happened, what the outcome was, what has happened since, and what new signals might have changed the picture.
And I particularly like your point about measuring how often a negotiator rejects the suggested reason for contact.
I’d actually see that as valuable feedback rather than failure.
If the system repeatedly gets the reason wrong, the system needs to learn that.
Ultimately, I don’t want PropFlow deciding who an agent should call.
I want it to give the agent a prioritised, evidence-backed shortlist and make the reasoning transparent:
Who. Why now. What happened previously. What changed. What to say next.
Then measure what happened.
Did it become a conversation? A valuation? An instruction? Was the recommendation rejected, and why?
That’s where I think this gets much more interesting than simply putting a score next to every CRM record.
This is the really interesting part of CRM data.
The value isn’t necessarily in having more records. Most established agents already have thousands.
The challenge is turning that history into something an agent can actually act on.
A database might contain previous valuations, withdrawn instructions, old buyers, past vendors and years of customer interactions.
But knowing that information exists isn’t the same as knowing who deserves a call today.
I think that’s where AI can become genuinely useful in estate agency — not simply generating content or automating another task, but analysing the signals already sitting in the CRM and helping agents identify where the next opportunity may be hiding.
Who to call. Why now. What to say.
That’s a much more interesting use of historic CRM data.