We’re looking for our first Lead Data Scientist: an independent technical lead who will check and challenge the new audience methodology on behalf of the whole OOH industry.
Our move to a new measurement approach, developed and run by Ipsos, is the biggest change to Route in years. It combines survey and mobility data with machine learning to simulate how people move around Great Britain and which Out of Home frames they pass and how likely they are to see them.
More sophisticated models bring a clear responsibility. Media owners and agencies fund Route jointly, and they trust our numbers because we are independent of any single interest. That independence has to extend to the methodology itself. We need our own expertise in-house, able to understand, test and, where needed, scrutinise how the models are built, so that Route can stand behind every number it publishes.
What the role covers
The Lead Data Scientist reports to Route’s Chief Executive Euan Mackay and has four key duties:
Scrutinise the methodology. Review the documentation and validation evidence as it evolves, and identify where it falls short of what Route would need to understand, test or improve the approach.
-
Set the bar for quality. Lead Route's input on acceptance tests and thresholds for each part of the methodology and each quarterly release, and represent Route in technical discussions, including our Technical Advisory Group.
-
Build independent checks. Benchmark outputs against census and mobility data, build simpler baseline models that show whether the complex ones earn their place, and trace how modelling choices carry through to published reach and frequency.
-
Translate for decision-makers. Advise Route's leadership and board on the strengths, risks and limitations of the methodology in terms a non-technical audience can act on, and give technical input to contract discussions.
The post-holder will work closely with our CIO Ian Wyatt and the data engineering team, and help build understanding of the methods across Route.
Essential criteria, skills and experience
We want a hands-on modeller with at least five years building machine learning and statistical models, who is as comfortable critiquing other people's work as building their own.
The essentials are deep learning experience, including sequence or generative models, and the judgement to know when a simpler model would do as well. You will need a track record of rigorous evaluation of generative or simulation models, practical survey statistics, and experience auditing models, often without access to the code. Just as important is the confidence to explain technical risk to a board and to push back on external experts.
Experience in synthetic populations, mobility and transport modelling, or geospatial and mobile data would be a strong bonus, as would a background in audience measurement or media. If you meet most of the essentials and some of the desirables, we would still like to hear from you.
Ready to apply?
The role is permanent and full-time, based in central London with hybrid working, and pays up to £100,000 depending on experience, with a strong pension and private medical cover. The full job description can be downloaded here.
Shortlisted candidates will complete an independent technical assessment set by a specialist outside Route, including a hands-on modelling exercise.
To apply, send a CV and covering letter to jobs@route.org.uk by 11 December.
