Pricing Domain Modernization for Dynamic Experimentation

The impact

Pricing modernized from fragile legacy logic to a centralized, reusable platform service. Modular experimentation enabled across Buy and Sell flows. Dynamic pricing strategies for high-demand events unblocked. Living roadmap and architecture artifacts left behind to sustain the work.

I led discovery and delivery for StubHub's pricing architecture modernization, turning siloed legacy pricing logic into a flexible platform that could support real-time experimentation, modular APIs, and future dynamic pricing. Leadership needed to see ROI early, in a three-week window, so the whole game was choosing the right first cut.

The constraint behind the constraint

StubHub's pricing tools were a liability, but the surface complaint, "the code is brittle," was not the real problem. Logic was scattered across internal tools, ops workflows, and legacy code with no documentation and no clear owner, and the downstream effect was the one that actually cost money: pricing strategy ran on instinct because the systems could not support anything better. The org was leaving dynamic-pricing revenue on the table not because nobody had the idea, but because the architecture could not hold an experiment.

I ran event storming, BORIS, and bounded-context mapping to get cross-functional agreement on what the current state truly was, not what the diagrams claimed. That surfaced POET, a core piece of pricing infrastructure, as both the critical dependency and the worst source of brittleness.

Three weeks buys one cut, so make it the right one

With only three weeks, the temptation is to map everything and ship nothing. The leverage move was the opposite: prioritize the POET refactor as the single early win that unlocked downstream work without breaking live buyer and seller flows, and sequence everything else behind it. From there I worked with engineering, pricing strategy, and platform teams to define a clean domain model and future-state APIs for Buy and Sell, mapped synchronous and asynchronous state-change messaging across critical domains, integrated Optimizely and pricing aggregators for experimentation, and built modular pricing object models that made A/B testing and price guidance tractable. All three OKRs landed in the window, plus a living roadmap the team could continue without me.

What this built

• A centralized pricing platform the engineering team could actually iterate on. Reusable APIs across Sell Flow, Buy Flow, and Auto Pricing logic replaced the legacy patchwork with something owned and maintainable.
• Real-time experimentation capability. Dynamic pricing for high-demand events, A/B testing across flows, and integrated metrics dashboards became possible for the first time.
• A living roadmap and architecture foundation. The strategy model, Miro diagrams, and phased migration plan gave StubHub a clear path forward, not just a snapshot of work done.

Sources

• Alberto Brandolini, EventStorming -- the modeling method (with BORIS bounded-context mapping built on it) used to map the pricing domain