What Geoteric tells us about the faults with SIC

There are many faults with SIC. In this post we’ve picked out one of the most interesting, and niche, companies from our dataset on artificial intelligence companies in the UK. Geoteric illustrates how these companies, to quote The Damned, “can’t be found, no way at all” using SIC.

They’re all but invisible in the official data because standardised industrial classification simply can’t cope with these types of companies. In many ways they defy categorisation.

Niche niche niche

Geoteric specialises in geological interpretation. They have created technology and AI that produce highly accurate data on geological structures. Geoteric’s customers use their technology on a variety of projects, from deepwater exploration, windfarm placement to well drilling. As their website says, ‘there is no doubt that AI has a significant future in the energy business, whether it be in oil and gas or the renewables sectors’.

Their marketing team gets an additional hat-tip from us for their line, “We’re really down to Earth. Really down to Earth”. Good stuff.

Exposing the faults in SIC

It’s the kind of company that standardised industrial classification can’t cope with. Indeed, when Geoteric registered at companies House it did so under SIC 58290 – ‘Other software publishing’. But, of course, that doesn’t really matter. Geoteric are successfully changing their industry regardless of the issues of industrial classification at Companies House.

It only matters when you want to analyse a sector, or when you want to find a specific type of company. You couldn’t find Geoteric using SIC. Would you search for a company like them under ‘Other software publishing’? Probably not. And you wouldn’t find them under the 30+ codes that other geology companies use, because they don’t adequately represent what Geoteric does. There isn’t even a code for AI, which is why we’ve created our AI RTIC.

SIC Codes are out of date, but that’s only part of the problem. It could be argued that the SIC Codes created in 2007 just need updating. But that misses the point. Hierarchical industrial classification won’t work these days. Innovation happens too quickly for static codes to be able to keep up.

A seismic shift is needed

Hierarchical classification breaks in the face of today’s economy. For example, is Geoteric a geology company using AI and tech or is it a tech company doing geology? Hierarchical classification systems like SIC depend on there being an answer to this question.

Something else is needed. We’re in the process of looking at different information architectures, including polyhierarchies and ontology design, to find a better way. We’ll be posting more of our emerging thoughts on ontologies and new data standards over the next few months as part of our commitment to working out loud.

We’ll also be posting on Twitter, and sending updates out on our monthly newsletter, so please subscribe if you’d like to hear about how we’re reclassifying industry.

About the author