We often talk about there not being enough skilled builders, bricklayers, electricians and plumbers to meet the demand for new housing in the UK. Less often do we speak about the significant skills shortages in surveying. Unfortunately, this is a reality confirmed by the Royal Institution of Chartered Surveyors.
Encouraging more young people as well as those looking to retrain to enter the profession is essential to meet demand, the trade body has said – more than once. Yet RICS also recognises that the nature of surveying services needs to do more to adapt to the changes seen in the wider economy. Technology, artifical intelligence (AI) and a better use of data is central to the development of our sector.
Not only does this require more people and younger people, crucially, we need to consider the changing needs in the valuations market and the skillsets that are required to meet them. All the way back in 2019, RICS published a report entitled Future of the profession.
In it, they recognised the major drivers of change in our industry, along with the areas that businesses were only just starting to think about at the time. They identified AI and the internet of things, better utilisation and management of data and decentralising traditional business models as thematic trends at the forefront of the industry’s minds.
Six years on and we are now living in this reality. ChatGPT, Google CoPilot and Microsoft Gemini’s AI tools are central to the majority of consumers’ lives around the world. The nature of how users interact online, search online and the speed at which they expect comprehensive information has completely transformed in a very short space of time.
Much as there has been better technology adoption in the valuations sector, we have not kept pace with the radical changes in technology’s capabilities. It is not surprising – surveyors work in a regulated market and interact with heavily regulated sectors in our value chain.
Data use comes with significant data protection considerations. The ability of generative AI tools in valuations is untested – where accurate and robust property risk assessment is critical to FTSE 100 blue chip companies’ financial resilience and their own capital valuations, relying on this type of tech at this stage is not yet practically feasible.
This doesn’t mean we can be complacent however. As part of the homebuying and remortgage value chain, surveyors have a responsibility to improve customer experiences just as much as lenders, advisers and conveyancers. Cost pressures and process efficiencies are ramping up all the time, particularly for mortgage business that is written at scale, skinny margins and largely using automated processes to keep costs down.
Valuations must keep pace with these needs. We are seeing a change in lenders’ appetite for automated valuation models in larger volumes of lower risk cases. The key here is in that word risk – AVMs have been used in the UK residential housing market for two decades, but the environment in which they exist has changed significantly over that time.
The most fundamental shift has been in the quantity of data available to feed into these models – some available openly, but also protected data sources have become much more diverse and the quality of that data is far more robust.
We’re on the cusp of changing the game in valuations – our AVM doesn’t rely on static sources of data, outdated data streams and aggregated risk assumptions. We have integrated a comprehensive range of open source data which is materially enriched by our own comprehensive proprietary data.
We have, at e.surv, created a property data lake of our own, fed into by our 500 valuers and millions of physical and desktop valuations. We can look at real-time data relating to non-standard construction and cladding exposure. Our surveyors on the ground add to this data lake every day, bringing in the nuance that outliers contribute to the overall context the model has to work from. We triage surveys based on real risks rather than the risk of risk.
A significant downside to generative AI, which is emerging on a ballooning scale at the moment, is that models rely on the data and information available online. They can correlate data from hundreds, thousands of sources and filter conclusions based on probability. But it is also highly prone to proliferating incorrect information, with the damaging effects of this growing exponentially at a terrifying rate.
Lenders cannot risk this when it comes to hundreds of billions of pounds of consumers’ and shareholders’ money. The integrity of AVMs must be verified and importantly the decisions and outputs they produce need to be explained and – that is what makes e.surv different; understanding, explaining and validating property risk decisions is what we excel at, whether data-driven or surveyor led.
But it is the choice that we offer our lender partners that sets us apart – from physical, building and condition surveys all the way through to data-based valuations. It’s up to you to scale and flex the services we provide to the changing needs of your business.
Steve Goodall is managing director at e.surv
