Predicting complaints and their causes to proactively improve service for Peabody residents

Peabody has a rich heritage as one of the UK’s longest-established not-for-profit housing associations. It is also one of the largest, currently managing 107,000 homes with 220,000 residents. Peabody’s priority is to support people and communities to flourish. This means providing the highest standards of service, including managing issues like repairs and complaints.

With such a wide portfolio, Peabody holds a wealth of data about its residents. That data can help the organisation to understand trends, predict certain issues and make sure that customer services levels meet its ambitions.

Red Olive was already working with housing association Catalyst when it merged with Peabody in 2022. Following a recommendation by Catalyst to Peabody, Red Olive helped Peabody prepare for compliance inspections. This led to conversations about how it could help in other business areas by applying data and analytics.

Developing a deeper understanding of complaints

Peabody felt that a focused project could help them improve in three areas:

  • Improve services to residents
  • Get the basics right first time
  • Deliver services more sustainably

Following a collaboration to design and develop tools for monitoring and improving customer data quality, Peabody asked Red Olive to work with them on developing a deeper understanding of customer complaints. The goal was to use data and analytics to understand what was driving complaints and how Peabody could better manage services and processes to improve life for residents.

“As this was the first data science project for the organisation, Red Olive played a pivotal role in demonstrating the technical aspects and complex methodology through clear and digestible communications to the business areas,” says Tracey Price, Director of Contact Centre & Complaints at Peabody. “This helped ensure that the business areas worked well with data science experts in delivering the project.”

The initial work took around three months. Red Olive investigated the business problem and designed a set of machine learning algorithms to identify residents who were more likely to complain, based on many factors including their experience of repairs and contractors, characteristics of their property itself, type of tenancy, financial considerations and so on. Further work focused on understanding the most prominent reasons why someone would complain.

Understanding these motivations, alongside geographical data, helps Peabody to see where issues may be more likely to arise and work proactively on services to support those residents. In testing based on understanding thousands of historical complaints over some years, the models accurately predicted who would be more likely to complain with suggestions on what that complaint would be about. While this does not guarantee that future complaints are predicted with the same level of accuracy, the models have given Peabody an improved picture on who is more likely to complain, and why.

housing industry infographic inside

First-of-a-kind project

The team has found both the data management and the associated analytics useful in connecting better with residents, allowing them to hear first-hand stories, and giving them a better understanding of common issues and frustrations.

It’s likely that this collaboration will move into further phases as Peabody’s data maturity level progresses, with the organisation using real time analytics to continue examining key elements of the business operations, such as the root causes of complaints, repairs volumes, and different tenancy types.

“This project is the first of its kind for Peabody,” says Ian McDermott, the organisation’s CEO. “It demonstrates our ambition to use innovative and advanced technologies to drive service improvement.

“Red Olive has been genuinely invested in the outcomes. Its ability to work in partnership with Peabody and bring its data expertise and practical knowledge of the housing sector has been helpful.”

Peabody is now also working with Red Olive to co-develop a data platform to deliver improved data accessibility and transparency. The platform will also give a consistent view of structured and unstructured data from a range of operational systems. Peabody expects this platform to further improve reporting and analytics, and the first output has already been delivered, focused on repairs contractor performance.

Jefferson Lynch, CEO at Red Olive, says: “Some of our other property clients have told us that the work we’ve done recently with them has really helped their customer engagement. Based on our experiences with organisations like Peabody, we have developed a complaints prediction and resolution software module, which has proved more than 80% accurate at predicting who would be more likely to complain with suggestions on what that complaint would be about. It identifies residents that are at high risk of complaining, and why, before they become so unhappy that they complain.”

“The work with Peabody involved large volumes of data, but not all organisations have the resources to train and build models at this scale. We built our module to make it accessible and affordable even for much smaller landlords. We can provide it as a fully hosted service, we just plug it into a client’s live data feeds, or we can install it inside their estate if that’s what they’d prefer. Because we’ve tested at scale, we’re confident that we can cope with situations where an organisation has relatively little historical data.”

“Reaching out to customers before a complaint has escalated, can improve customer relations as well as reduce business cost and influence efficiency for the landlords we are working with. We are extremely excited about our new solution as we can see the impact it is having on those organisations who are implementing it.”

Get more value from your data

Contact Us