Senior Product Manager, AI Experimentation Platform

CSC Generation•Cottonwood Heights, UT

11h•Hybrid

About The Position

Backcountry is moving from manual, low-volume A/B testing to an always-learning agentic platform that designs, launches, monitors, and optimizes experiments across its ecommerce sites. The Senior Product Manager will own the platform from the ground up, setting its roadmap, deciding where AI agents can act autonomously and where human oversight is needed, and driving measurable improvements in conversion and revenue. The role reports to the Vice President of Product and UX and is based in Cottonwood Heights, Utah, with a hybrid schedule requiring three days per week in the office. The role has end-to-end responsibility in a lean team. Within six to twelve months, success means having a working platform that runs trusted experiments across multiple sites, increasing testing velocity without sacrificing statistical rigor, and achieving revenue and conversion improvements that can be traced to platform-run experiments. The position partners with engineering, data science, design, analytics, merchandising, and marketing to build platform infrastructure, establish experimentation standards, and make testing self-serve. Backcountry offers paid time off policies, a 401(k) match, medical, dental, vision and supplemental policies, and employee discounts across its portfolio of brands.

Requirements

At least 5 years of product management experience, including ownership of a platform, data, or internal tooling product.
Hands-on experience running or building A/B testing and experimentation programs at scale in ecommerce or other high-traffic digital products.
Strong statistical grounding in experiment design, including power and sample size calculation, significance, peeking, multiple comparisons, and guardrail metrics.
Hands-on experience building with AI and LLMs, including agentic workflows, prompt design, tool integrations, and evaluating AI output quality.
Working knowledge of experimentation infrastructure, including feature flagging, randomization, event tracking, and warehouse-native metrics.
Strong analytical skills and hands-on use of SQL, GA4, BigQuery, Looker, or similar tools.
Demonstrated measurable impact on conversion, revenue, or testing speed.
Excellent communication skills and ability to influence engineering, data, and business stakeholders.
Provide a valid LinkedIn profile URL in the application, or give a credible reason for not having one and supply alternative evidence of professional background.
Attend a mandatory in-person interview before an offer; candidates unable or unwilling to attend will be removed from consideration.
Employment is contingent on a background check, including verification of prior employment dates.

Nice To Haves

Experience with experimentation platforms such as Optimizely, Statsig, Eppo, GrowthBook, LaunchDarkly, or an in-house equivalent.
Experience with adaptive experimentation, including multi-armed or contextual bandits, Bayesian methods, or variance reduction techniques such as CUPED.
Experience testing AI- or ML-driven experiences such as search, recommendations, or personalization.
Experience running experimentation across multiple brands or storefronts on a shared platform.
Background in outdoor, specialty, or multi-brand ecommerce.

Responsibilities

Own the vision, roadmap, and prioritization for the agentic experimentation platform, including deciding what to build first and what to buy versus build.
Define how AI agents participate throughout the experiment lifecycle, from hypothesis generation and variant creation to test setup and QA, live monitoring, and ship-or-rollback decisions.
Establish safety guardrails for autonomous testing, including statistical standards, sample ratio mismatch and anomaly checks, revenue and margin thresholds, approval workflows, and automatic rollback.
Build adaptive optimization using multi-armed and contextual bandits so traffic can shift toward winning experiences while tests are running.
Partner with engineering and data science on assignment and bucketing, feature flags, event tracking, the metrics layer, and a statistics engine built on GA4 and BigQuery.
Define and instrument platform success metrics, including test velocity, win rate, revenue impact, and result trustworthiness.
Make experimentation self-serve for merchandising, marketing, UX, and engineering teams.
Create standards and playbooks for metric definitions, test design, and readouts so results are consistent, comparable, and trusted across sites.
Partner with engineering, design, analytics, merchandising, and marketing to turn business goals into a shipped platform and a culture of testing.