Subscribe
About

Behavioural-based rental app sparks interest after debut

Simnikiwe Mzekandaba
By Simnikiwe Mzekandaba, IT in government editor
Johannesburg, 06 Apr 2021
Zabeth Venter, CEO and co-founder of Averly.
Zabeth Venter, CEO and co-founder of Averly.

Following its debut in the local market last August, some 4 000 tenants have signed up to use locally-created rental application Averly.

This is according to CEO and co-founder Zabeth Venter, telling ITWeb that “quite a few” agents and landlords have signed up and are administrating their business on Averly.

“We can definitely see that it [Averly] sparked an immense interest in the market,” she says. “We decided to do a slow take-up…because we wanted to see whether the performance is up to our standard. There’s a lot of interest.”

Founded by a husband and wife team in the property industry, Averly is a progressive web app (PWA) that combines technology and behavioural analysis to help rental agents and landlords to screen prospective tenants.

It uses machine learning and neuroscience technology to verify credentials, analyse track records, and aggregate factual data in order to predict personal behaviour, says the company.

Venter states Averly is aimed at creating collaboration between the tenant, agent and landlord.

“No more sending contracts via WhatsApp or e-mail, everything is on the app – the communications happens on the app.

“The benefit of a PWA is that you do not have limited functionality on the app. What you have on your Web site is exactly what you have when you work on your mobile phone. It automatically scales to whatever phone you are using.”

Digital revolution

Averly digitises the entire rental application process. A link is sent to prospective tenants to create a profile and fill out the survey of 22 questions. Tenants can also upload all important and verified documentation onto the platform that would normally have been scanned and e-mailed to the rental agent.

Venter says they noticed the tenant, agent and landlord relationship was sometimes very “difficult” to manage. “What we’ve built is a platform that is embedded with emotion AI [artificial intelligence] that measures what people say versus how they feel.”

The behaviour management is done via a score. “People understand things like an Uber or Google score – they are used to it. Users get an Averly score and that will go up and down depending on how interactions go in the rental space.

“There is this enormous expectation gap in the rental industry. You tell me that you will be a good tenant but then you don’t pay your rent. What the app does is give information right at the start of the rental process about the parties you want to enter into a contract with.

“This is to prepare the agent or landlord for what might come down the line. We realised that we had to create the app so that admin takes less time and relationship-building is a priority.”

Averly also helps with managing property maintenance, as agents can use it to build a comprehensive move-in snag list during inspections, upload photographic evidence, and indicate where maintenance work is needed.

It points out that tenants have access to the same list, leaving no room for nasty surprises when they move in or out at the end of their lease period.

The company explains that if maintenance is needed during the rental period, tenants can indicate it on the same list. This will trigger a notification to the agent to arrange the relevant fixes. Rental agencies can also use this functionality to monitor the service their agents provide.

To use the app, agents and landlords pay a monthly subscription of R30 per property. “It works on a sliding scale; the more properties one has on Averly, the cheaper it becomes. We really wanted to make it affordable. Tenants get the app for free at this stage.”

Minding the bias

On whether such an app has the potential to create some form of bias, Venter notes they are very aware of the inherent bias that sits in data.

“A lot of people think the bias is created, but how you go about working with data is to understand the context and whether the data being used is fit for that context.

“When we created our emotion AI, that was the very first step that we made sure we get right. We understood the context and we understood all the potential types of bias that can be hidden in our data.

“We designed the process in such a way that we can easily identify and correct the process.

“If someone says their model doesn’t have inherent bias at all, they either do not know it or don’t understand it yet. What’s important is how you manage it and we definitely manage it on an active basis.”

Venter believes all three parties – the agent, landlord and tenant – stand to benefit from using the app.

“Currently, the agent and landlord have the benefit of the tenant Averly score when completing an application for a property.

“Later on during the year, we want to introduce an agent and landlord score, which will give the tenant the same information about the agent and landlord when applying for property.”

Share