Webinar: Data Analysis and Data Science: The new frontiers of real estate research and decision-making, 11th November 2020
The art of data science
For many in the real estate world, data science may conjure up visions of geeky boffins slaving over impenetrable computer programmes, but this SPR webinar suggested that it – or at least its application – is as much an art as a science and that real estate professionals ignore its growing influence at our peril.
Of the four speakers, John Affleck of CoStar, who was beamed in from the US, possibly waxed most lyrical. He suggested that writing the code required to make sense of his organisation’s voluminous store of property-level data was like penning poetry – it’s all a question of finding exactly the right words and placing them to perfection. To do this you need to know what you’re trying to achieve and what your client needs to know, whether it’s a rental forecast or a story of vacancy rates.
Whatever the recent lurid headlines about algorithms consigning surveyors to history, Dan Hughes of Alpha Property Insight agreed with Affleck that humans would continue to have a major role in communicating the meaning of information and in deciding how to respond to it in the real world. Although the quantum of data potentially available to real estate professionals is now growing exponentially, there are many steps still needed to standardise disparate sources and resolve ethical issues like privacy before the true potential of ‘big data’ can be realised for the industry.
The importance of this subject to the audience was confirmed when moderator Joanna Turner of Canada Life opened an online poll asking how important it is for real estate organisations to have a digital strategy and a data science department, with most answering ‘very’, although Hughes retorted that strategy was much more crucial than having internal staff, as smaller organisations would probably do better to buy in externally created data.
There were many questions from the audience. Responding to Ruth Hollies, CBRE, who asked how data users could be sure of its accuracy, Samantha Kempe of IMMO Capital, one of the panellists, said that her organisation spends a huge amount of time cleaning and questioning data. A key challenge is finding automated ways to identify anomalies, but there will always need to be human involvement to make sure the data can really support the uses to which it is being put. Later in the discussion, Kempe emphasised the great potential of data for residential property, the sector her organisation is mainly involved with. Here there tend to be many more transactions and data points than for commercial real estate, but their exploitation has so far been limited.
Answering a question from SPR Fellow Yolande Barnes about the most valuable uses of data science in the investment process, Dominic Silman of LaSalle IM suggested that the relatively long underwriting period for commercial real estate makes it possible to undertake detailed pre-acquisition analysis, particularly of the geo-spatial kind. This can draw on a wealth of data, for example on competing enterprises in a locality. But he also stressed that as most real estate decisions were of a long-term nature and involve large sums of capital, the human role in interpreting data remains crucial.