AI moves from hype to hard returns in property

Artificial intelligence is moving beyond experimentation in South African real estate, with property companies deploying technology to cut costs, accelerate leasing, improve decisions and unlock new investment opportunities.

• AI and automation are already cutting leasing administration, processing thousands of leases and freeing property teams from repetitive work.
• Construction, property management, data analysis and affordable-housing investment are emerging as major areas where AI can improve returns.
• The message from property leaders is clear: digitise first, start small, measure efficiencies and keep humans involved in critical decisions.

Artificial intelligence may be the property industry's favourite technology buzzword, but the real question for investors and operators is becoming much simpler: where is it actually delivering returns?

That was at the heart of the panel discussion “Embracing Innovation: Transforming the Property Sector through Technology and AI” at the IHS Affordable Housing Conference at The Houghton Hotel in Johannesburg.

Moderated by Matt Marshall, Partner and Co-Founder of Re-Dimension Capital, the discussion brought together Wayne Burger, Co-Founder of SA Proptech; Nic Foce, Managing Director of Foce Properties; Nico Papanicolau, CIO of TUHF Capital; and Peter Stainton, Executive Head of Property Management at Eris Property Group.

Marshall framed the shift as one from digital workflows to something much bigger: property companies have accumulated vast quantities of data, but AI is increasingly making it possible to bring that information together and turn it into intelligence.

The discussion showed that AI's immediate value may not lie in replacing property professionals, but in making buildings cheaper to develop and operate, processing information faster and allowing people to spend more time making decisions that actually matter.

Peter Stainton: Technology must make property operations work

‍For Peter Stainton of Eris Property Group, the clearest gains are already appearing in high-volume property management and leasing.

Student accommodation provides an extreme test case. Stainton said the business can process 5,000 to 6,000 leases in January alone - volumes that would simply not be manageable without technology.

Digital systems now allow leases to be processed, stored and retrieved efficiently, with the next opportunity moving from the income side towards reducing inefficiencies in operating expenses.

His key points included:

• Automate high-volume leasing: technology is essential when thousands of leases must be processed within weeks.
• Attack operating costs next: once leasing is digitised, facilities management and operating expenses become the next efficiency frontier.
• Augment people rather than automatically replace them: AI can make property professionals better informed and more consistent.
• Choose technology partners carefully: not every solution needs to be developed internally, and failed technology should not be retained simply because money has already been spent.

Stainton's message was pragmatic: technology should solve an identifiable property-management problem, not be adopted because AI happens to be fashionable.

Nic Foce: Attack the biggest cost - development

‍Nic Foce of Foce Properties shifted the conversation upstream to where some of the biggest property costs occur: development and construction.

While owners may fight for percentage-point improvements in operating margins to achieve 8%, 9% or 10% yields, Foce argued that debt and construction account for the overwhelming share of development cost.

That makes construction efficiency a much larger lever. He cited simple AI applications such as recording site inspections and automatically converting them into minutes and instructions for teams. The technology may appear basic, but multiplied across projects and employees, small time savings can become significant.

‍

His philosophy is straightforward: “Let’s get rid of the boring stuff, let’s focus on the people.”

‍

His key points:

• Target time, not technology: Foce challenges employees to find hours every week that AI can save.
• Construction is ripe for disruption: reducing development lead times could materially reduce financing and project costs.
• Automate predictable work: if a computer can reliably perform a repetitive process, allow it to do so.
• Keep human problems human: sensitive tenant circumstances and relationship-driven decisions still require people.
• Use property data commercially: utilities, energy and other building services can potentially create additional income streams.

The lesson is important for developers: AI doesn't need to transform the entire construction industry tomorrow to generate value today. It needs to remove enough wasted hours and costs to improve feasibility.

Wayne Burger: Digitise before you AI

‍For Wayne Burger of SA Proptech, one of the industry's biggest mistakes is attempting to leap directly into AI while underlying systems remain fragmented.

Property companies frequently operate multiple spreadsheets and legacy platforms that cannot communicate with each other. Burger argued that modern technology stacks need open APIs, allowing systems to connect and information to flow into central dashboards.

And there is already measurable evidence of the benefits. He cited a multifamily client that digitised its lead-to-lease process and achieved a 40% saving in lead administration workflow and a 70% saving in stationery costs.

His key points:

• Digitise first: AI cannot perform optimally on fragmented, paper-based processes.
• Build connected systems: open APIs and centralised data should be minimum requirements.
• Accept that technology projects sometimes fail: don't keep building around a bad system because of sunk costs.
• Empower employees to experiment: people who understand existing processes are often best positioned to identify where AI can save time.
• Protect the environment: AI introduces significant cybersecurity and governance risks.

Burger strongly advocated keeping a “human in the middle” while AI systems mature. Automating an incorrect process can simply produce mistakes faster and at greater scale.

Nico Papanicolau: From automation to AI agents

‍Perhaps the most advanced AI implementation discussed came from Nico Papanicolau of TUHF Capital. TUHF has developed what Papanicolau described as an agentic technology stack, with AI agents being implemented across approximately 56 organisational roles.

Instead of employees manually moving between Salesforce, SAP, Power BI, email and document systems to assemble information, AI agents can interrogate multiple systems, consolidate the information and return an answer.

The objective is to drive the administrative burden on highly skilled professionals towards zero and allow them to concentrate on higher-value work.

His key points:

• Connect organisational data: emails, documents and operating platforms collectively contain intelligence traditional databases miss.
• Use agents for routine decisions: AI can increasingly handle high-volume micro-decisions based on defined evidence and rules.
• Move beyond cost savings: the next stage is asking AI to identify new revenue and profit opportunities.
• Use AI to understand unstructured data: particularly valuable in affordable housing, where conventional datasets may be incomplete.
• Build security in from the start: sophisticated AI deployment requires governance and guardrails, not security added afterwards.

Papanicolau's contribution pointed towards the next phase: AI moving from assisting employees to becoming an operating layer across the property business.

Where AI is making the biggest impact

‍Across the panel, five areas emerged where technology is already beginning to move the needle: Leasing and property management. Construction and development. Operating-cost optimisation. Data-driven investment and tenant decisions. And employee productivity.

But the bigger opportunity could be affordability. Better data, faster feasibility analysis, lower administrative costs and more precise decision-making could ultimately help developers and funders determine how to deliver housing to markets that conventional property models struggle to serve.

Marshall suggested that the real opportunity is to use technology to make development sufficiently precise that the industry can serve a broader market, including households currently trapped in an informal market where, as he put it, “the poor pay the most.”

The winners won't be those with the most AI

‍The strongest takeaway from the IHS discussion was not that every property company needs a grand AI strategy tomorrow.

It was almost the opposite. Start with the problem. Digitise the process. Connect the data. Automate repetitive work. Measure the saving. Then introduce AI where it can make a better decision or create additional value.

Foce's challenge to his own employees perhaps summed it up best: how many hours can you save this week?

Because ultimately the property industry's AI race will not be won by the company using the most sophisticated terminology. It will be won by the businesses that can turn technology into shorter development timelines, lower operating costs, faster leasing, better decisions and ultimately stronger property returns.

AI is moving from the property industry's innovation agenda to its investment agenda and that is where it starts to matter.

‍