Senior Product Manager - Catalog

At Salla, we are building the digital backbone of commerce in the region, empowering thousands of merchants to start, grow, and scale their businesses. We turn what's complex into something simple, powerful, and scalable. Catalog is the layer underneath everything. Storefronts, search, recommendations, landing pages, and every channel a merchant sells through all read from the same product data. When it is good, every surface above it gets better at once. When it is thin or wrong, nothing built on top can fix it. Two things make this exciting:. First, Salla does not build the catalog. Merchants do, and often not even the merchant: a freelancer, operations staff, an outsourced agency. Every surface we build inherits what those hands produce. So the question is not how to clean the data. It is how you change what someone produces when they do not work for you and mostly want to finish and move on. Second, no two categories need the same product record. What a fashion merchant needs is not what a pharmacy, a grocer, or a gold merchant needs. The easy answer is to build each category separately, and it does not scale. You will drive impact across five areas - The Catalog Engine and Category Setup (one engine, built once, with what each category needs from a product record expressed as configuration rather than code, so supporting one more category is fast and repeatable) - Product Data Input (creation, bulk import, editing, and everything that happens at the moment someone types or uploads) - Structure and Meaning (taxonomy, attributes, variants, identifiers, normalization, and duplicates) - AI Enrichment (extracting and generating product data from what merchants already have, with a clear position on confidence, review, and correction) - Downstream Fitness (how this data actually performs where it is consumed: search results, storefronts, recommendations, and external channel feeds) Responsibilities: - Own the catalog strategy end to end: sequence it into quarters, defend it in the room, and hold it accountable to the numbers it promised. - Own the roadmap for how product data gets created and improved on Salla, prioritized on whether the data works downstream rather than on how many fields are filled. - Design for the people who actually do the work, including the ones who do not own the store, and make the good path the fastest path rather than a longer one enforced by rules. - Keep a new category, a configuration change rather than an engineering project, and be able to show the difference in effort between the first one and the next. - Put AI to work where it earns trust: decide what we generate outright, what we suggest for approval, and what we never touch, with a stated confidence policy and a loop that learns from every correction a merchant makes. - Define what catalog quality means at Salla and instrument it honestly, tied to the surfaces that consume the data rather than to completeness for its own sake. - Own how product data performs in external selling channels, where the requirements are strict and the rejections are visible. - Decide how much structure we impose and where merchants stay free, then defend that line against pressure from both directions. - Lead 2 to 3 product managers: give them real ownership, coach their framing and craft, run a reliable cadence, and raise the quality bar of what the group ships. - Work as one system with the teams that depend on this data: search and discovery, storefronts, vendors, marketplace, and data. Requirements - Six or more years in product management, including work on a self-serve platform where users created their own content or data and you were accountable for the quality of what they produced. - A number you owned and moved on that data: completeness that meant something, accuracy, time to publish, adoption of a data tool, or the performance of that data somewhere downstream. - Evidence that you changed what users produced at the moment they produced it, rather than cleaning it up afterwards in a pipeline. We will ask which one it was. - Platform judgment: you have built something once and made it serve many cases through configuration, and you can explain where you drew the line between what was shared and what was specific, and what that choice cost you. - AI in production for structured data: extraction, classification, generation, or matching. You can talk concretely about confidence thresholds, error rates, what you let through automatically, and what you did when it was wrong at scale. - Literacy in how product data is actually structured: taxonomies, attribute schemas, variants, identifiers, matching and deduplication, and the strict requirements external selling channels impose. - You have coached product managers, formally or not, and you are ready to lead a small team without stepping away from the craft yourself. - You own outcomes rather than roadmaps, you move quickly through ambiguity, and you build clarity where others see nois

Six or more years in product management, including work on a self-serve platform where users created their own content or data and you were accountable for the quality of what they produced. A number you owned and moved on that data: completeness that meant something, accuracy, time to publish, adoption of a data tool, or the performance of that data somewhere downstream. Evidence that you changed what users produced at the moment they produced it, rather than cleaning it up afterwards in a pipeline. We will ask which one it was. Platform judgment: you have built something once and made it serve many cases through configuration, and you can explain where you drew the line between what was shared and what was specific, and what that choice cost you. AI in production for structured data: extraction, classification, generation, or matching. You can talk concretely about confidence thresholds, error rates, what you let through automatically, and what you did when it was wrong at scale. Literacy in how product data is actually structured: taxonomies, attribute schemas, variants, identifiers, matching and deduplication, and the strict requirements external selling channels impose. You have coached product managers, formally or not, and you are ready to lead a small team without stepping away from the craft yourself. You own outcomes rather than roadmaps, you move quickly through ambiguity, and you build clarity where others see noise.

Own the catalog strategy end to end: sequence it into quarters, defend it in the room, and hold it accountable to the numbers it promised. Own the roadmap for how product data gets created and improved on Salla, prioritized on whether the data works downstream rather than on how many fields are filled. Design for the people who actually do the work, including the ones who do not own the store, and make the good path the fastest path rather than a longer one enforced by rules. Keep a new category, a configuration change rather than an engineering project, and be able to show the difference in effort between the first one and the next. Put AI to work where it earns trust: decide what we generate outright, what we suggest for approval, and what we never touch, with a stated confidence policy and a loop that learns from every correction a merchant makes. Define what catalog quality means at Salla and instrument it honestly, tied to the surfaces that consume the data rather than to completeness for its own sake. Own how product data performs in external selling channels, where the requirements are strict and the rejections are visible. Decide how much structure we impose and where merchants stay free, then defend that line against pressure from both directions. Lead 2 to 3 product managers: give them real ownership, coach their framing and craft, run a reliable cadence, and raise the quality bar of what the group ships. Work as one system with the teams that depend on this data: search and discovery, storefronts, vendors, marketplace, and data.