Opportunity at
DeliverooStaff Machine Learning Engineer - Ads Bidding
This is a full-time, on-site Staff Machine Learning Engineer role in London within the advertising machine learning group. The group develops and maintains the models that…
Career Tools
About This Role
This is a full-time, on-site Staff Machine Learning Engineer role in London within the advertising machine learning group. The group develops and maintains the models that drive Deliveroo's ad offering, matching shoppers with relevant promotions and supporting merchants, grocers and FMCG brands in growing sales through the app. The London contingent counts roughly eleven ML engineers across ads and promotions, and it works with a US counterpart team, with meetings…
Job Description
This is a full-time, on-site Staff Machine Learning Engineer role in London within the advertising machine learning group. The group develops and maintains the models that drive Deliveroo's ad offering, matching shoppers with relevant promotions and supporting merchants, grocers and FMCG brands in growing sales through the app. The London contingent counts roughly eleven ML engineers across ads and promotions, and it works with a US counterpart team, with meetings capped at 4pm UK time.
The main undertaking is shifting ad pricing from cost-per-click to cost-per-order, meaning merchants pay only when an advert produces an order. The posting presents this as a frontier problem with little precedent, and the method is still being debated. The engineer would take responsibility for designing and carrying out the change, including whether the outward model stays order-based while clicks are tuned internally, or whether a blend of the two is chosen.
Further duties cover the bidding product itself, spanning goal-driven automated bidding, target cost-per-order, optimised cost-per-click and target return on ad spend, plus yield management connected to ROS targeting. Day-to-day coordination with the US group keeps the approach consistent across markets, and the engineer sets the technical course for the domain.
Suitable candidates understand how advertising economics and auction-based pricing work, and have hands-on familiarity with the bidding strategies above. The posting asks for seven or more years of production Python work, sound grounding in ranking, recommendation or bidding systems, and a preference for simple designs that produce measurable business gains.
Responsibilities
- Take ownership of designing and delivering the transition from cost-per-click to cost-per-order.
- Work out with the team how an order-based external model should combine with internal click optimisation.
- Build and refine the bidding surface, including goal-driven automated bidding and target return on ad spend.
- Take responsibility for yield management linked to ROS targeting.
- Keep the approach aligned with the US team across markets through daily contact.
- Set the technical direction for the bidding domain.
Requirements
- Grounding in the economics of advertising, auction design or marketplace pricing.
- Direct, hands-on work with optimised cost-per-click, target cost-per-order, automated bidding and target ROAS.
- Self-driven leadership, able to set direction where no path is established.
- Seven-plus years of shipping Python to production, with firm machine learning foundations in ranking, recommendation or bidding systems.
- A leaning toward straightforward designs and toward shipping work that moves business numbers.
- Clear cross-team communication, including with Product, Commercial and the broader DoorDash group.
Core Skills
Benefits
- Competitive salary and equity options.
Frequently Asked Questions
Answers are based only on the employer’s listing; where it doesn’t say, neither do we.
Where is this role based?
The role is based in London and is listed as on site. The team works closely with a US counterpart.
What type of employment is offered?
This is a full-time position.
Is a salary stated in the posting?
The posting does not state a specific salary. It mentions competitive salary and equity options.
What is the core technical challenge?
The role centres on moving ads from cost-per-click to cost-per-order bidding, which the posting describes as an open problem.
Sample Interview Questions
AI-generated questions tailored to this specific role — a preview of the full practice set.