Senior Product Analyst
Individual-contributor analytics lead at PDQ, which runs A/B tests in e-commerce merchants' checkouts. Owns experiment design and readouts, product and SaaS metrics (adoption, churn, MRR/NRR/LTV), and builds AI-driven automation as skills in their MAXX AI analyst agent. Heavy SQL/Python, Tel Aviv office 4 days a week.
- Compensation
- Not disclosed
- Location
- Tel Aviv-Jaffa, Tel Aviv District
- Work setup
- Hybrid
- Seniority
- Senior
- Employment
- Not specified
- Experience
- 6+ years
- Category
- Data analytics
- Posted
Skills
Must have
- 6+ years in product or business analytics, in e-commerce or B2B SaaS
- Hands-on A/B testing: hypothesis, sample size and power, duration, randomization unit, sample ratio mismatch, segment cuts
- Product thinking: turn a vague question into a metric, a test, or a clear decision not to measure
- AI-first builder: uses AI agents daily and has built something others use (agent, skill, or automated analysis); verifies output against data
- Strong SQL and comfort in Python
- Presents to CFO, Sales VP or CSM in their language and defends numbers from the data
- Self-driven and curious; finds the question before anyone asks it
- Tel Aviv office four days a week; full professional English
Nice to have
- E-commerce or DTC domain experience, on the product or merchant side
- Hands-on Snowflake experience
- Agentic analytics: semantic models an agent queries, MCP tools, guardrails against invented numbers
- Understanding of frequentist and Bayesian experimentation (sequential testing, CUPED, multiple-comparison control)
- Churn, LTV or pLTV modeling
Full description
About the role
PDQ runs A/B tests in our merchants’ checkouts. You own experimentation and product analytics, and every result you report must hold up when someone challenges it.
You lead without a team. You pick the problems, ship the work yourself, and set the bar for the analysts around you.
You think AI-first, and you build. When an analysis repeats, you turn it into a skill in MAXX, our AI analyst agent. When a team still works by hand, you automate it. You shape how PDQ uses AI, in the product and across the company. Nobody assigns you this part of the job. You take it.
What you’ll own
- Experimentation. Test design, sizing and readouts for all merchants, and a repository of what we learn from wins and losses.
- AI in analytics. New MAXX skills, and automation of the analysis work teams still do by hand.
- Product analytics. How PDQ measures adoption, activation and impact, and the definitions behind them.
- A CRO framework for 200+ active merchants. Segmentation, opportunity scoring and expected impact that MAXX and CSMs run per merchant.
- Churn and expansion. The loop from churn signal to intervention to outcome, and a per-merchant ROI scorecard for Sales and CS.
- SaaS metrics and win/loss. MRR, NRR and LTV, the monthly close with Finance, and a quarterly readout on why we win and lose deals.
Requirements
- 6+ years in product or business analytics, in e-commerce or B2B SaaS. If you come from SaaS, you love online shopping.
- Hands-on A/B testing (we check it live in the interview): hypothesis, sample size and power, duration, randomization unit, sample ratio mismatch, and segment cuts. You know when a winning segment is noise.
- Product thinking. You turn a vague question into a metric, a test, or a clear "we should not measure this".
- AI-first builder. You use AI agents every day, and you built something with them that other people use: an agent, a skill, or an automated analysis. You check the output against the data and defend every number without the AI.
- Strong SQL and comfort in Python.
- You present to a CFO, a Sales VP or a CSM in their language, and when challenged you answer from the data.
- Self-driven and curious. You find the question before anyone asks it.
- Tel Aviv office four days a week. Full professional English.
Advantages
- E-commerce or DTC domain experience, on the product or merchant side.
- Hands-on Snowflake experience.
- Agentic analytics: semantic models an agent queries, MCP tools, and guardrails that stop an agent from inventing a number.
- Good understanding of frequentist and Bayesian experimentation (sequential testing, CUPED, multiple-comparison control).