Product Engineer — Monetization & Commerce Data in London

See something that should exist? Build it. See a better way? Do it. At Terrific, every role directly impacts how we grow from product to revenue to customer experience. We’re building the next generation of social commerce through live shopping powered by real-time data and AI. And we believe customer data should belong to the brands who earn it — not third-party platforms.

See data that should be worth money? Make it worth money. See a number nobody can explain? Go find out why. At Terrific, we're building the next generation of social commerce—helping brands and publishers turn content, creators, live experiences and AI into commerce they own. Our technology brings shoppable video, live shopping, interactive timelines, polls and product carousels directly onto a merchant's website or app. Now we're looking for a Product Engineer to turn what those experiences already know about a shopper into a monetization product—and to decide what that product should be.

THE OPPORTUNITY

This is not a role where you'll inherit a roadmap. You'll write it. Terrific sits on something most advertising companies cannot get. On a single customer's domain we see the video someone watched, the poll they voted in, the product they tapped, the coupon they used, what went into the cart and what they actually bought—first-party, consented, tied to a real catalog and a real checkout. Most advertising systems have one half of that and model the other half. Nobody has turned it into a product yet. That's this role. You'll build the layer that lets a retailer sell advertising inside their own site to the brands they already stock, and lets a publisher prove what their audience is worth in transactions rather than impressions. There is no ad server today. No auction. No campaign model. No advertiser interface. It is a blank page, and you get to decide what goes on it.

WHAT YOU'LL BUILD

• First, the thing only we can build: measurement. Impression, view, engagement, product interaction, add to cart, checkout, purchase. Client-side and server-side. One canonical event model, identity resolution, consent, and reporting a brand manager can read without an analyst.
• Then the thing that makes money: on-site retail media. Sponsored placements on inventory we already control—timeline, shoppable video, live—sold to brands the retailer already stocks. No auction required. No ad server required.
• After that, connectivity. Campaign and audience APIs into external buying platforms, and eventually programmatic supply through Google Ad Manager and Prebid. We expect to rent more of the advertising stack than we build. You'll help decide which parts.

WHAT YOU'LL OWN

• Writing the spec. You get a problem and an account, not a ticket.
• Designing the event model across every Terrific surface—video, live, timeline, polls, chat, cart, checkout.
• Identity resolution and consent, built so our customers keep their data structurally and not only contractually.
• Attribution: connecting advertising exposure to transactions, and being able to defend the method to a customer in one sentence.
• Reconciliation between third-party reporting and our own logs, which never match.
• The advertiser-facing surface—campaign setup, pacing, delivery, reporting.
• Sponsored placement delivery and ranking on Terrific inventory.
• Build-versus-rent recommendations, including which providers you'd shortlist.
• Instrumenting what you ship and watching it get used.
• Deciding what not to build this year, and saying so out loud.

You won't be expected to know advertising technology on day one. You will be expected to form an opinion, test it cheaply, and change it when the evidence says so.

WHO YOU ARE

You're an engineer who thinks in products. You've shipped data systems that carried real volume, not prototypes. You're comfortable with Node.js and TypeScript, strong SQL, and event pipelines—BigQuery and Pub/Sub, or close enough that you'll be fluent in a week. You know what's actually buildable, how long it will take, and when to simplify scope to get something real into production faster. That last part is the job, not a compromise. You care about the product and not only the code—which means you'll push back when what you've been asked for doesn't make sense. You can sit with a retailer's analytics lead, work out what they actually need to see, and go build it without a product manager in between. And you're honest about numbers. In advertising, almost everyone's numbers are slightly wrong and most people don't say so. We'd rather be the ones who can explain the difference.

HOW YOU'LL USE AI

Terrific is AI-native by default. We don't treat AI as a separate initiative or an occasional productivity tool. We use it as part of how we investigate, build, document and improve our work. In this role, you'll have the freedom to use AI to:

• Prototype a data model in an afternoon instead of a sprint.
• Generate and test transformation and reconciliation logic.
• Explore a customer's catalog, feed or event stream quickly.
• Draft the specs and documentation you'd otherwise postpone.
• Build internal tools that let commercial teams answer their own data questions.
• Monitor pipelines and catch discrepancies before a customer does.

We're interested in what you can build with AI—not whether you use a particular tool. Use it to move significantly faster. Don't outsource your judgment to it.

WHAT WE'RE LOOKING FOR

• At least five years building production data or backend systems.
• Node.js and TypeScript.
• Strong SQL.
• Real event-pipeline experience—BigQuery, Pub/Sub, or close equivalents.
• Experience owning something end to end: you decided what it should do, then built it.
• Ability to design a data model someone else can still understand a year later.
• Clear written communication. You'll write specs other people build from.
• Good judgment about effort, risk and commercial value.
• Comfort working independently within a distributed, international team.
• Professional fluency in English.

A strong advantage:

• You've joined advertising exposure data to transaction data before. Attribution, identity resolution, clean rooms, or conversion APIs.
• Retail media, commerce media, or publisher monetization.
• Analytics or reporting products used by non-technical customers.
• Spanish, Portuguese or French for our LATAM and EMEA accounts.

You do not need to have built an SSP, and you don't need ad-operations history. If you've built serious data products and you can hold a product argument, the advertising domain is learnable and we'll teach it.

HOW WE HIRE

No take-home project. About four hours of your time in total.

• A call with the hiring manager. Thirty minutes.
• A pairing session in our actual codebase, with one of our engineers. One hour. Not a whiteboard, not an algorithm puzzle.
• A product judgment conversation. We give you a real scenario: a retailer with five million monthly sessions, a product catalog, purchase data, Terrific inventory and Google Ad Manager, who wants brands to buy sponsored placements and wants purchases attributed to the advertising. We want to hear what you'd build first, what you'd refuse to build in year one, and where you'd cut scope.
• A technical design conversation. One hour.

We'll tell you what we're assessing before each round, and we'll give you feedback either way.

SUCCESS LOOKS LIKE

In your first months, you will:

• Ship an event and identity layer that covers every Terrific surface.
• Produce the first attributed revenue number a customer actually trusts.
• Put a working sponsored-placement product in front of one retailer's brand partners.
• Give our commercial teams numbers they can sell on.
• Recommend what we build and what we rent—and be right often enough that we stop asking twice.
• Leave behind a data model the next three engineers don't want to replace.

Success is not shipping every item on a roadmap. Success is a retailer being able to say "advertising on my own site produced this much revenue"—and prove it.

WHY THIS ROLE IS DIFFERENT

Most engineering roles in advertising begin with a ticket and end when the ticket is closed. This one begins with a blank page. You'll decide what the product is, not only how it gets implemented. You'll work directly with customers, with the people making commercial decisions, and inside a company that already has the data—just not yet the product. The advertising industry spends enormous effort modeling what Terrific observes directly. You'll be the one who turns the observation into revenue.

REQUIRED

• Based in Europe, or within a compatible European time zone.
• Available for a full-time engagement.
• Able to work remotely with teammates and customers across global time zones.
• Professional fluency in English.
• At least five years of hands-on production engineering.
• Experience with event pipelines, data modeling and analytics infrastructure.

Terrific operates as a global company, and English is our primary business language. Interviews, documentation and regular internal communication will be conducted in English.

Required:

• Deep experience in at least one of Terrific’s three commercial pillars — eCommerce, Marketing, or Media — and enough fluency across the other two to contribute credibly from day 1.

Terrific operates as a global company, and English is our primary business language. As such, all roles require fluency in written and spoken English.

Be part of the change that’s quietly reshaping how people discover and buy online. Passionate about social media and e-commerce? Come build what’s next at Terrific.

Please note: We review every application carefully. While we may not be able to respond to everyone, we will reach out to candidates whose experience aligns with what we’re building. We may use artificial intelligence (AI) tools to support parts of the hiring process, such as reviewing applications, analyzing resumes, or assessing responses and identifying potential inconsistencies or verification signals in application materials based on available information. These tools assist our recruitment team but do not replace human judgment. Final hiring decisions are ultimately made by humans. If you would like more information about how your data is processed, please contact us.