The Cabinet Office and Frontier Economics Research the Size of the Geospatial Market with Glass.AI.

The Cabinet Office within the UK Government set up the Geospatial Commission in 2018 to unlock the economic, social and environmental opportunities offered by location data. It recognised that the capabilities offered by geospatial data and location-based insights were no longer confined to the periphery of the UK economy. However, one of the challenges that the geospatial market presents is that it doesn’t operate like a traditional market sector, but has emerged as an ecosystem of organisations offering a diverse range of products and services across a variety of different markets. This complexity makes it difficult to measure the size, value and impact of geospatial data.

The Geospatial Commission engaged with Frontier Economics, a London-based economics consultancy, to perform a detailed economic study of the geospatial data market in the UK to feed into their own policy and development recommendations in the Geospatial Commission report on ‘Enhancing the UK’s Geospatial Ecosystem’. Recognising the difficulty of identifying the companies within the geospatial market, Frontier Economics collaborated with glass.ai to identify geospatial companies from the large dataset of UK companies that our AI has collated from crawling the UK business web and using our language understanding technology to read the descriptions and other related content of businesses from their own websites.

Distribution of Geospatial Businesses over the UK

The difficulty of identifying geospatial companies determined to mean “organisations which could not deliver their product or service without geospatial data”, led to an iterative approach to building the market definition. Simply including all companies who had a mention of a particular geospatial keyword in their description would have led to over-estimating the market size. For example, drones manufactured for use in leisure or media activities should not be included in market, but those used for agricultural surveys should be included. In the general case, these sorts of distinctions can be very hard to capture. By building a language model around relevant geospatial terminology, and capturing rules for the context in which this terminology should be used, we were able to quickly produce small samples of results for review. These would be manually checked, and the results fed back to improve the model. The iterative process was able to continue until a high level of quality of results was obtained. In the end, around 2,000 companies for which the supply of geospatial data is a core activity were identified. Extrapolating from the available turnover and employee data matched to these companies, the total turnover for the market was determined to exceed £6 billion with over 115,000 people employed — highlighting the scale of the opportunity that the UK has to develop this market.

Hard to define markets, such as Geospatial, are a common problem when trying to determine the scale of an opportunity or activity. Often this will be because of a service (e.g. call centre), technology (e.g. artificial intelligence) or skillset (e.g. animation) that cuts across many traditional industries. However, the Geospatial Data Market identification further demonstrates that by applying scalable web reading technology and using the human-in-the-loop to review and give feedback over iterative cycles, complex collections of companies can be gathered to match almost any criteria. If you have your own market that you want to uncover then you can get in touch here.

To read more about the development of the Geospatial Market Data study and the recommendations for nurturing the geospatial economy in the UK, you can visit the Government’s Enhancing the UK’s Geospatial Ecosystem report page here.

Previous
Previous

KTN Taps Into Glass.AI to Identify Emerging Technology Companies and Their Decision-Makers.

Next
Next

The ONS Uses Glass.AI to Understand the Trading Conditions of Businesses in the Middle of a Pandemic.