Ex-Amazon entrepreneur founds ‘Tinder for property’
The process of finding a new rental property is broken – and artificial intelligence can fix it.
That is the claim made by former Amazon entrepreneur Paul Bennett, co-founder of PerchPeek.
The platform offers a user experience similar to that of Tinder by asking users to swipe left or right when searching a new place to live.
But the platform’s true ingenuity lies in its approach to using data. While browsing, users are asked their preference on books, films and television.
The approach might seem ‘outside the box’ – but it is based on Bennett’s working knowledge of the most successful website on the planet.
A former business development manager for Amazon’s Treasure Trucks service, Bennett – like the eCommerce giant’s CEO Jeff Bezos – knew that customer experience is key to a successful website.
“[At Amazon] the entire focus is on machine learning to recommend things to customers,” he told BusinessCloud.
“Every time you go on Amazon it’s a completely dynamic experience – the homepage is always unique.”
Yet after helping his mum manage her small property portfolio in Guildford – and struggling to find a property himself in Oxford – he realised the property management experience was ‘laborious and inefficient’.
“On Rightmove and Zoopla, every time you go back it’s the same database,” he explained.
ENTREPRENEUR BUILDS ‘BREXIT-PROOF’ ARCHITECTURE DISRUPTER – CLICK BELOW
Bennett clarifies that neither he nor the company know exactly how a taste in music affects a person’s taste in property, but he believes that a correlation will emerge in the data which allows the right landlord and tenant to meet.
“We’re in the process of gathering data at the moment. We don’t have enough data to effectively recommend at a much higher level yet. The challenge for us gathering enough good quality data,” he said.
“The premise is the same [as Amazon] – similar customers want similar things.”
He and his team of six in the UK – with another five people working remotely – are trying to lower the average 35 enquiries which occur before the right tenant moves in.
They have already beaten this average with their 7,000 listings of London and Guildford properties – across 148 suppliers – but are aiming to bring the average down further still.
“We want to drive that conversion lower and lower, because a one in 35 conversion is awful,” Bennett said.
“There is so much wastage in the industry and that’s why so many people recognise renting to be a stressful process.”
The company has to date helped more than 150 people find a property and recently reached 100,000 searches through the platform.
It has also received backing from PiLabs and Damien Marmion, the former CEO of AXA Global Health.
As well as renter’s interest in entertainment, the company also collects more obvious data points about its users, including the area they are interested in and price range.
This data become more useful over time as its users, which are primarily 18 to 35 years old – or ‘generation rent’ – continue to move around.
“We know that there are a large proportion of people, a third of millennials, will be renting their entire lives. We want to be the platform that they do that through,” said Bennett.
It also hopes to move down the supply chain, offering not just a listing platform but a website which can handle references and contracts.
The start-up’s decision to only list properties in London and Guildford is a conscious choice of ‘quality over quantity’, Bennett says, but the plan is to expand its marketing budget to get more tenants on to the platform.
Bennett’s ambition is to start collecting data about properties and renters in all of the UK’s big cities within 12 months.
Armed with years of invaluable data about the type of property each user prefers, it hopes to become the de-facto platform when making a move.
Bennett said he can imagine a future where landlords go straight to PerchPeek instead of their estate agent.
It is also hoping into Europe, with a focus on countries that have an even higher proportion of renters.