Real estate is ahead of the curve in proving a use-case for AI

Over the past few weeks in the US there has been a growing debate about whether the enormous investment ‘Big Tech’ is making in AI infrastructure will ultimately deliver the returns expected. Billions of dollars are being poured into new data centres, compute clusters and model-development pipelines, but the timelines for monetisation are long, usage patterns are still evolving and the path to reliable revenue remains unclear. The market is starting to ask whether there is a business case that can justify the scale of spending.

What is interesting is that the commercial property sector, so often a laggard in tech adoption, has already established a clear use case for AI and we are already seeing the automation of sales collection and analysis impacting how retail and leisure assets are managed and ultimately valued.

For decades retail property has suffered from a lack of transparency around sales performance. Many landlords still rely on quarterly or even annual reporting, often delivered manually and in inconsistent formats. By the time the data arrives, the trading environment may already have shifted – and with it, the true performance of the tenant or the asset. This lag has real consequences in making leasing decisions slower and less informed, introducing unnecessary risk into rent negotiation processes, and, perhaps most importantly, undermining the accuracy of asset valuations.

Automated systems can now collect sales data in close to real time, clean it, structure it and analyse it without the delays and gaps that can hold back decision-making. When this data is processed through AI models, it becomes far more than raw numbers, revealing patterns related to seasonality, promotional cycles, footfall dynamics, tenant mix changes and external economic conditions. Investors, valuers and asset managers can see how different parts of a destination are performing day by day – and even predict how they are likely to perform in the weeks and months ahead.

This greater level of transparency is very attractive for institutional investors seeking long-term, secure income. The shift to turnover-based leases made retail property more unpredictable, in part because opaque or infrequent data made it difficult for investors to assess risk with confidence. With continuous, AI-enhanced visibility analysts can differentiate between cyclical fluctuations and structural trends, cashflow forecasts become more robust and valuations reflect trading conditions and performance. I believe this will fundamentally change the risk profile of these assets and open the door to investors who may previously have viewed retail as too volatile.

AI can be a game-changer, but only once large volumes of data have been collated and analysed, systems have been given time to learn and optimise with human monitoring. Retail destinations generate highly patterned behaviour – weekly rhythms, seasonal peaks, predictable responses to events and promotional activity. When AI models are trained on several years of detailed, consistently collected sales data, their predictive accuracy improves sharply.

They will have a deeper understanding of what is driving performance which in turn will inform leasing strategies and capital-investment. Even day-to-day activities such as marketing campaigns, community events and pop-ups can be modelled and evaluated based on their measurable effect on trading performance. The value of embracing AI-powered data collection will compound over time.

Those who delay adoption will find themselves at an increasing disadvantage. The information gap will widen year by year and in a data-driven market, slower insight means slower action. As valuations become more tightly linked to live performance data, owners without long-term datasets will struggle to provide the same level of confidence to investors.

In an environment where Big Tech is still searching for clear and reliable commercial returns from its massive AI bets, real estate is emerging as a sector with a measurable and compelling use case. Like the first wave of proptech innovation a decade below, landlords not only need to identify the problem and the fix they also need to select the right tech solution that is robust, efficient and can be integrated into a wider operating platform. If they get this right, significant value can be unlocked now, but even more so in the future.