Retail Data Analyst

Profitmind is a retail analytics SaaS company that turns competitive and customer data into agent-driven insights, helping retailers make faster, sharper merchandising and pricing decisions. Based in Pittsburgh and backed by a recent strategic investment from Accenture, we are scaling our agentic AI platform for some of the world’s largest retailers.

The Role

We’re hiring a Retail Data Analyst to own whether what we show a customer is right — end to end: agreeing how every metric is calculated before a customer’s data arrives, proving our computed numbers reconcile to their own reporting, and checking that the insights and recommendations built on those numbers make business sense before a customer ever sees them.

This is a retail domain role, not an infrastructure one — you won’t build the pipelines (Data Engineering) or the forecasting, elasticity, and optimization models (ML Engineering). You own the layer between them: the definitions, the reconciliation, and the judgment about whether an output is sane. We’re hiring retail judgment first — five or more years of retail data analytics — but the role is hands-on: you’ll query the warehouse yourself rather than describe the problem to someone who can.

You’ll partner closely with product, customer success, data engineering, and ML. Today this work is absorbed by six people across engineering — roughly seventy hours a week between them, none of them hired to do it.

What You’ll Do

• Agree the metrics before data arrives. Work through a business requirements questionnaire with each customer and turn their answers into a written, customer-specific definition of every metric — fiscal calendar, comparable-store definition, category hierarchy, and treatment of returns, markdowns, and fees — before a single file lands.
• Prove the numbers reconcile. Compare what we compute against a report the customer’s own business users read, and drive the two to agreement — hold the line on that source. When figures disagree (sales, units, AUR, profit, margin, discount rate), find a named cause at the category level, across last week, MTD, QTD, and YTD.
• Own comparable-store reporting. Define comparable, new, closed, and total store populations for each customer, and confirm every metric computes correctly for each — that the parts add up to the total, and that point-in-time measures aren’t split by store group as if they were a comparison.
• Pressure-test the output, not just the inputs. Review agent-generated insights and recommendations before customers see them, challenge results that look directionally wrong, and trace a wrong number to its real cause — missing data, a mapping error, or a definition never actually agreed on.
• Own the KPI library and document the boundaries. Build and maintain the standard reference for metric definitions and calculation logic;
  where a customer’s definition departs from it, record what differs and why. Document what the platform handles well and where its limits are, so customer success can set accurate expectations.
• Send findings back with evidence, and re-check after every release. Turn findings into structured feedback for product and engineering — what’s wrong, what it should be, the evidence for both. After each release, confirm the numbers that mattered still reconcile.
• Be the sign-off before insights ship. Bring your evidence to the end-of-onboarding review and either attest that the data and output are trustworthy, or stop the launch.
• Use — and improve — our automation. Much of this work is automated (requirements drafting, field mapping, data quality checks, reconciliation reporting). Run it, read its output critically, and tell us where it falls short.

What We’re Looking For

• Retail data analytics depth — 5+ years, required. Hands-on experience with retail data, inside a retailer or serving one (merchandising, pricing, planning, finance, or BI). Fluent in comparable-store definitions, retail fiscal calendars (4-5-4, 13-period, calendar-year), markdown/clearance vs. regular price, returns treatment, AUR, GM%, weeks of supply, sell-through, and open-to-buy. You should hear “comp sales were up 9%” and immediately ask which store population, which calendar, gross or net of returns.
• AI-assisted working. Comfortable folding AI tools into daily work to pull, check, and compare data faster — with the judgment to catch when the answer is wrong.
• Able to get to the answer in the data yourself. Comfortable querying the warehouse — joins, aggregation across grains — including with AI assistance, enough to reproduce a disputed number without handing it to an engineer.
• Able to judge an output, not only a number. Enough grasp of demand forecasting, price elasticity, and markdown optimization to say whether a merchant would believe a recommendation, and to tell a model problem from a definition problem.
• Reconciliation as a discipline. A track record of tying computed numbers to an authoritative source and closing gaps line by line (retail, finance, FP&A, revenue management, or BI) — comfortable explaining a variance to the person who owns the number.
• Clear communicator. Fluent in Excel/Sheets for variance comparisons;
  writes metric definitions, findings, and limitations that other teams can act on without a meeting;
  treats “same number, different definition” as the leading explanation for a variance.
• Customer-facing and self-directed. Can run a working session with a customer’s merchandising or finance team and hold a line on data quality without turning it into a fight;
  comfortable owning ambiguous problems end to end.

Nice to Have

• Experience on the retailer side — merchandise planning, pricing, allocation, or finance
• Python (pandas or equivalent) for validating, comparing, and reshaping files
• Experience onboarding enterprise SaaS customer data, or running data quality/profiling tooling
• Hands-on with a cloud warehouse or lakehouse (Databricks/Delta Lake, Snowflake, BigQuery, Redshift, or similar)
• Experience with AI or model-based products, and with BI tooling (Power BI, Tableau, Looker)

Location & Reporting

Hybrid - Pittsburgh, PA. Reporting line: TBD. This role works in close partnership with both engineering and product.

Why This Role

Every insight we ship rests on a customer believing our numbers. This role makes that systematic: you define the metrics with the customer, prove they reconcile, judge whether what we built on them makes sense, and own that standard from the first requirements call through every release afterward.

Profitmind is an equal opportunity employer. We welcome applicants of all backgrounds.