How a Five-Person Estate Agency Went From Monthly Guesswork to Real-Time Intelligence

They knew they wanted to open a second branch. They even knew the number: £500,000 in revenue. What they did not know was how close they actually were, or what was standing in the way.

This is a small London estate agency — five people, covering sales, lettings and valuations. Like most independent agencies, they were busy. But busy and clear are not the same thing. KPIs were tracked manually, sometimes on paper. Monthly meetings were spent piecing together numbers rather than acting on them. Nobody had a reliable view of where leads were converting, which channels were working, or where deals were quietly slipping away.

The decision making was happening. It was just happening without the right information.

Starting With the Real Problem

Before building anything, TUSTRA ran a full operational audit.

The goal was to understand how the business actually worked, not how it was supposed to work. What we found was that the core issue was not a lack of data. It was data quality. Records were inconsistent. Fields were missing. The same information was entered differently by different people on different days.

Building any kind of analytics on top of that would have produced unreliable results. So we started there.

We worked with the team to clean and standardise their existing data, introduced clear processes for how information should be entered going forward, and created simple guidelines to help the team maintain data health consistently. Once the team were on board and the quality was holding, we had a foundation worth building on.

Building the Intelligence Layer

With a year of clean, reliable data available, we used AI and analytics to calculate KPIs across every part of the business — sales, valuations and property stock.

A selection of what the dashboard tracks:

• Pipeline health — total valuations, live pipeline count and value, completed sales revenue

• Momentum indicators — new instructions added, valuations in the last 30 days, new sales value

• Performance metrics — completion rates, average time to exchange, withdrawal rates

• Ageing and risk signals — stock older than 90 days, live sales exceeding 120 days, oldest scheduled valuation

• Data quality scores — listings with incomplete data, valuations with missing price fields

These are not vanity metrics. Each one connects to a real business decision: where to focus, what to fix, and what is trending in the wrong direction.

From Numbers to Intelligence

The first version of the dashboard gave them visibility. The second version gave them foresight.

We built a layer of forecasting and anomaly detection on top of the existing data. Rather than simply presenting numbers, the system now identifies patterns that suggest something is about to go wrong — a sale at risk of falling through, a valuation backlog building, a conversion rate dropping before it becomes a problem.

When the system detects a signal, it surfaces a recommendation. The team knows what to look at and what to do about it, before it is too late to act.

At TUSTRA, this is what we mean by an intelligence layer. Not a dashboard full of numbers. A system that watches your business, tells you what matters, and gives you time to respond.

Updated 25/01/2026