Tahiti North is run by Robert Antolin, an operator who has sat on both sides of the Amazon table. Inside Amazon, as a Vendor Manager, he owned the full P&L for a major category: growing sales, protecting margin, and negotiating directly with vendors and Amazon's own retail teams. That seat is where the firm's read on Amazon comes from: how a category team actually decides to grow a vendor, throttle it, or stop buying from it.
After leaving Amazon, he moved to the other side of the table, advising brands as a Vendor Central expert. Tahiti North runs three practices: AI Transformation, eCommerce Strategy, and Fractional Executive Services, anchored by Amazon Vendor Central margin recovery for 1P brands.
Amazon · Sr. Vendor Manager, Furniture · 2018
Owning the full P&L for a major category meant growing sales and protecting profitability at the same time, the trade-off most vendors feel from the other side. Through pricing, terms, and annual negotiations, the category beat its 2018 sales goal by $28.1M (+17.7%), and across his tenure contribution margin improved by roughly 910 basis points.
Amazon · NA Retail catalog · 2016 to 2019
Two initiatives, a catalog automation and an algorithm-defect fix later adopted across Amazon's NA Retail org, restored ordering on SKUs that had been silently suppressed. The lesson drives this firm's recovery work: platform systems fail quietly, at scale, and the money goes to whoever reads the data closely enough to notice.
App in the Air · Chief Operating Officer · 2024
As commercial lead, redesigned pricing, subscriptions, and channel mix around what the P&L could prove rather than what the roadmap promised. Ancillary revenue grew 562% year over year at a 51.8% lower customer acquisition cost.
Tahiti North is operator-led, not agency-led. Every engagement is run personally by the same operator who owns the history above, not handed off to an account team once the contract is signed. Methods are productized: named playbooks and a recovery engine behind the margin recovery audit and the rest of the service line, not open-ended hours. They sharpen on a benchmarks flywheel, where every audit and every dispute feeds proprietary calibration data into the next one.
The same systems discipline shows up beyond Amazon: production LLM automation built and run personally at App in the Air (2022 to 2025) saved 5,900+ hours a year at under 2% error.
Engagements start with a working session on your numbers: an AI-readiness question, a monetization problem, a Vendor Central deduction backlog, or a leadership gap a fractional executive could fill.
Get in touch