Quantifying how predictive software reduces evictions for social landlords

A large-scale, independent study set more than 1.1 million social housing tenancies managed with rent-arrears prediction software against over 2 million managed without it. It found that software users cut evictions due to arrears by 37.8 per cent over three years, against a 13.3 per cent reduction for non-users. The same analysis linked the software to a 29.6 per cent fall in non-Universal-Credit arrears, worth an estimated £600,000 per 10,000 properties over two years, and to a roughly 11.5 per cent drop in the number of tenants in arrears before that trend levelled off. For the software provider, the study turned a plausible vendor claim into sector-wide, independently evidenced proof.

The problem. Social landlords have spent the past decade absorbing rising rent arrears and mounting eviction risk, driven by welfare reform and, in particular, the staggered rollout of Universal Credit, which changed how and when much housing-benefit income reaches tenants and landlords. Software providers whose products claim to reduce arrears and evictions in that environment face a credibility gap: cash-constrained landlords, housing boards and finance committees need more than a vendor's own case studies before committing budget to a system, and a claim that software "reduces evictions" carries little weight unless tested against a comparable group of landlords who did not adopt it. The provider at the centre of this study needed exactly that: externally credible, statistically defensible evidence, at a scale large enough to withstand scrutiny from boards, funders and regulators.

The method. QuantSpark carried out what is believed to be the most extensive quantitative study of its kind in the sector: more than 1.1 million tenancies managed with the software set against more than two million managed without it, tracked over a multi-year window from 2015 to 2018. Rather than a single before-and-after snapshot, the design split the tenancy base into a software-user cohort and a non-user cohort and followed both through the same multi-year period, so the results captured a sustained trend rather than a one-off dip. Four outcome metrics were isolated for separate measurement: evictions due to arrears, total arrears, non-Universal-Credit arrears specifically, and the number of tenants in debt. Isolating non-Universal-Credit arrears mattered because it stripped out the confounding effect of welfare-reform timing, letting the analysis attribute the remaining movement more confidently to the software itself.

The workflow. The analytical pipeline ran in five stages: assembling and cleaning tenancy-level data across both cohorts for the 2015 to 2018 window; segmenting outcomes into the four metrics above; running the comparative statistical analysis cohort against cohort, metric against metric; translating the resulting percentage differences into a monetary indicator scaled to a standard portfolio size, per 10,000 properties, so landlords of different sizes could compare like with like; and packaging the findings as an independent, sector-facing evidence base rather than an internal vendor report. The subject of the study, the landlords' own rent-arrears prediction software, sat at the centre of the comparison, with statistical and comparative cohort analysis tooling providing the methodology and the monetisation calculation layered on top.

The value. The headline result, a 37.8 per cent reduction in evictions due to arrears against a 13.3 per cent reduction for non-users, gives the provider a comparative claim, not just an absolute one: users of the software achieved a materially larger reduction in evictions due to arrears than landlords who did not adopt it, over the same three-year window. The £600,000-per-10,000-properties figure was built the same way as the other headline numbers: not as a one-off saving, but as an indicative monetary translation of a percentage-point arrears reduction (1.6 percentage points over 24 months, a 29.6 per cent relative fall) scaled to a common portfolio size, so landlords managing more or fewer than 10,000 properties can scale the estimate to their own stock rather than read it as a fixed sum. The roughly 11.5 per cent fall in tenants in arrears is reported as a trend that stabilised rather than one that compounded indefinitely, an honest signal that the benefit plateaus rather than growing without limit.

For a market where landlords are wary of vendor-supplied efficacy claims, an independent study spanning more than 1.1 million tenancies with the software and over 2 million without it, across multiple years, turned "our software works" into a statistically grounded, sector-benchmarked claim the provider could put in front of boards, funders and regulators.