From Aerospace to Art, Appraisal Bureau’s new Chief Technology Officer

Chief Technology Officer at Appraisal Bureau, Scott Strong, speaks to art market journalist Riah Pryor, about strategic use of data

You’ve worked across a broad range of companies and sectors, from personal finance [with Previsio] to cloud cost analysis [with CloudZero], and now with Appraisal Bureau, which specializes in the valuation of Fine Art and Digital Assets. What has your work across these roles had in common?

As a trained aerospace engineer, I was taught to systematically break down problems into their component parts, making it easier to understand. After experiencing business school, I gained an appreciation for the environments in which companies thrive. Combined, these skills have enabled me to enter a business or industry with no prior knowledge and understand what questions to ask and who might have the best answers. 

Having worked in so many industries now (from investing to fraud detection), I have learned to quickly absorb industry specific knowledge by speaking with subject matter experts and translate that knowledge into actionable data insights and machine learning products. While it doesn’t always lead to innovation using machine learning, I have found that the best way to contribute is to immerse myself in the business’s environment.

There is a lot of discussion about the opportunities and challenges machine learning could pose to the art sector. What do you feel are the biggest changes it could have on the sector?

I like to think of machine learning as a tool, not a solution. What that means in practice is that when trying to solve a problem, you want to use the most appropriate tool for the job. Sometimes that’s machine learning, and other times it’s not. If you try to force machine learning as a solution, that’s when things can get messy.

In the art sector I see many opportunities to leverage machine learning and technology more generally to improve the experience of owning and managing an art portfolio. For example, if you can demonstrate to an insurance company that your machine learning enabled solution is faster and more accurate than traditional options in delivering art valuations, they will quickly be converted. The trick is finding situations in the art sector that leverage the strengths of machine learning and avoiding situations where it’s a forced solution.

This certainly makes sense when it comes to larger companies who may have the capacity and resources to explore machine learning, but what advice would you give to smaller companies or individuals who may wish to experiment with boosting their use of data - are there any quick wins?

Many businesses have data that can provide unique insights into their customers and/or their industry. The key here is getting specific about what you as a company want to do with your data. Data science and machine learning require data to be organized, but you can’t organize your data unless you have goals in mind!

Sometimes you can do this on your own, but many times it makes sense to bring in a data expert for a consultation on how your business can best leverage its data. While there are some upfront costs, this approach will ensure you don’t waste time or money in the long run.