Head of Experimentation
About the Job:
Feature management and experimentation have converged into a single market, and the buying dynamic at the top has shifted. Engineering teams are no longer the sole evaluator — data scientists and data-focused PMs now carry equal weight on the largest deals. The bar for statistical depth, warehouse ergonomics, and experiment-first workflows is rising quickly.
In traditional experimentation we have built the foundation: a trusted runtime control plane, a growing experimentation engine, and early warehouse-native capabilities. We are winning lower-maturity buyers at healthy rates. We are not yet consistently winning the most sophisticated data organizations. Closing that gap is the job.
In AI experimentation, we have an early lead: the AI-native tooling category has invested in evaluation and conceded production experimentation, and we already have the primitives (statistical significance, multi-armed bandits, experiment-aware guardrails) that no AI-native competitor ships. Extending that lead is the other half of the job.
This leader will own whether LaunchDarkly becomes the definitive experimentation platform in an AI-accelerated world.
Responsibilities:
-
Own the Experimentation pillar. Direct leadership of the Product team. Partner with Engineering and Design counterparts in a triad model. Accountable for the pillar's strategy, roadmap delivery, and commercial outcomes. Make the investment case across the in-product experimentation experience, the warehouse-native analysis layer, and the infrastructure that scales them.
-
Make experimentation the measurement layer of the AI SDLC. Partner with our AI product, observability, and core feature management leaders to productize the capabilities we already have as AI-native primitives. Build a closed loop from offline evaluation through production experiments, to automatic promotion and rollback, to a self-improving feedback loop for agents.
-
Win the high-maturity buyer. Earn the technical confidence of senior data scientists and data-focused PMs. Decide what statistical depth, warehouse coverage, and experiment-first workflow capabilities are non-negotiable, and get them shipped on a timeline that wins pivotal reference deals.
-
Make warehouse-native a weapon. Expand coverage across major data warehouses and query layers. Deliver parity on analysis-only mode, variance reduction, ratio and percentile metrics, exposure validation, and arbitrary-window analysis.
-
Operate a high-performing function. Run a disciplined roadmap, ship predictably against quarterly commitments, drive AI-assisted engineering productivity inside the org, and hire where gaps exist.
-
Be the external face of the category. Credibly represent the product with Data scientists, PMs, experimenters, analysts, and partners. Translate the strategy to the field and equip sales to win head-to-head.
How you'll be measured:
- Win rate on experimentation-involved deals, especially head-to-head competitive evaluations — step change in the first year, sustained improvement thereafter.
- Reference-grade customers at the top of the maturity curve, including named strategic logos.
- Monthly active customers and active-account ARR growth against plan.
- Experimentation attach rate on new and expansion enterprise deals.
- Engineering throughput — roadmap delivery velocity and AI-assisted development adoption inside the function.
Qualifications:
- Senior product leader (GM, VP, or equivalent) with a track record of owning a product line that competes on statistical rigor and data infrastructure.
- Deep, operator-level fluency in experimentation methodology: causal inference, variance reduction, ratio metrics, sequential testing, exposure design, multi-armed bandits, and composite/multi-objective metrics — and the realities of running these at scale against production data warehouses and against non-deterministic systems where output variance, not just user variance, drives sample