Senior Product Manager - Technical

Deep Sync is the industry leader in deterministic identity and AI-powered data solutions. Leveraging our 35-year foundation of compiling direct mail-grade datasets, Deep Sync develops data-first technologies that power marketing, measurement, AI, and business intelligence for agencies and brands with our fully deterministic identity graph, privacy-first audiences, comprehensive data solutions, and integrations with leading platforms and cloud providers.

Marketing depends on identifying the audience correctly. Deep Sync supplies the identity data that makes that possible, across offline and online channels, in both consumer and business marketing.

You will own the product surface on which a customer selects an audience, purchases it, and delivers it into the channel where the campaign runs, built on the platform now under development. The role also covers new product development: taking an idea from a hypothesis to a paid offering, and discontinuing the ones that do not justify further investment.

The products we build have to be ready for an AI first world. Customers increasingly reach data through agents and natural language rather than through a screen, which means what we ship has to serve a machine as readily as a person.

The position calls for a senior product manager fluent in advertising and in data, and equally at ease with AI. You should already understand what a match rate is worth to a buyer, the limits of lookalike modelling, and why one individual appears as separate records in separate systems, you should be able to go into the data and verify such questions yourself, and you should be building with current AI tools rather than reading about them.

We run a flat organisation and a fast product development cadence. The position is based in Kirkland, WA.

• Surface ownership. Own an area of the product from the problem through to the outcome: the case for doing the work, the requirements engineering builds from, and the usage it shows six months after release.
• Requirements and written decisions. We work in documents. A requirements document states the problem, the decision taken, what is in scope and what is not, and how success will be measured. Presentations are for reporting, not for deciding.
• Scope discipline. A rebuild invites reproducing every feature that exists today. Establish what is actually used, make the case for retiring the rest, and hold that position in review.
• Release management. Select the customers who receive each release first, define the test that closes a milestone, and report the result as it stands.
• New product development
• Opportunity identification. Attributes that have never been packaged, audiences that have never been priced, and questions customers are currently paying somebody else to answer.
• Early validation. A test with two or three customers, a measure agreed in advance, and a short written finding. Evidence before a long build rather than after it.
• Applied AI. Determine where a model or an agent belongs in the product and where it does not, specify how its output will be evaluated, and demonstrate the idea with a working prototype rather than a description of one. Design for consumption by machines as well as by people: predictable interfaces, described data, and output that a system downstream can act on without a person in between.
• Progression or discontinuation. Both are acceptable outcomes. An idea that remains a proposal for three quarters is not.

• Customer contact. The sales, delivery and client facing teams hold most of the knowledge about what fails, and customers hold the rest. Time spent with them is part of the role.
• Consumer and business…