Swish
Website:
justswish.in
Job details:
At Swish, we’re reimagining food delivery by combining speed, freshness, and consistency. Our platform delivers high-quality meals, snacks, and beverages in under 10 minutes, turning everyday eating into a seamless experience. We recently raised a $38M Series B to accelerate our expansion across multiple cities, invest in kitchen infrastructure, and strengthen our supply chain. We’re building a fast-growing, high-ownership company focused on redefining how India eats every day.
This is a critical analytics hire in the organization and one of the most consequential roles in the team. You are not joining an established data organisation with clean pipelines and documented metrics. You are building it. The dashboards, the metric definitions, the event taxonomy, the reporting cadence — all of this starts with you.
This is also a team management role. You will hire and develop two Growth Analysts beneath you, and you are responsible for their output quality and growth.
What you will own
- Metric framework design — Define and document every metric in the KPI framework — from the north star (orders per day) through L1 drivers (new user orders, repeat user orders, AOV, GPPO) to L2 health metrics (cohort retention, frequency buckets, segment migration, GPPO decomposition). Every metric needs a precise definition, a calculation method, an owner, and a review cadence.
- Data infrastructure — Set up and own the analytics stack. Configure the product analytics tool (Mixpanel or Amplitude), connect CRM data feeds, build the pipeline from the transactional database to the reporting layer.
- Core dashboards — Build and maintain the dashboards that run the business. Daily OPD tracker. Weekly growth dashboard across all pods. Cohort retention curves by acquisition week, channel, and city. Frequency bucket migration matrix. GPPO decomposition by store and category. These are not reports that go out once — they are live instruments that the team uses every day.
- Cohort analysis — Build and maintain the cohort model that underpins every retention and lifecycle decision. D1, D7, D30 retention by signup cohort, by acquisition channel, by city, by store. Revenue retention curves separate from order retention curves. This is the most important analytical output in the function — it drives CRM journey design, acquisition channel allocation, and payback period modelling.
- A/B test framework — Define the experimentation standard for the team. What constitutes a valid experiment, how to calculate required sample sizes, when a test has reached statistical significance, and how to read results correctly. Run the statistical reads on all active experiments. Prevent the team from acting on noise.
- Team management — Hire, onboard, and develop a team of Growth Analysts. Define their scope clearly so they are not stepping on each other. Review their work before it goes to pod leads. Build their analytical capability over time.
Metrics you will be held to
- Dashboard accuracy — Zero wrong numbers in any live dashboard. Every figure has a documented calculation and a named owner.
- Instrumentation coverage — All key product events are tracked correctly before features launch. No retroactive data loss.
- Experiment read quality — A/B tests read at correct significance thresholds. No decisions made on inconclusive data.
What we are looking for
Must have
- 4–5 years in a growth, product, or business analytics role — ecommerce, quick commerce, or food delivery strongly preferred.
- Has designed a metrics framework from scratch — not inherited one. Can explain the logic behind every metric definition they have built.
- Strong SQL — can write cohort queries, window functions, and complex joins without help. This is non-negotiable at this level.
- Hands-on experience with a product analytics tool — Mixpanel, Amplitude, or equivalent. Has built funnels, cohort analyses, and retention curves, not just read them.
- Has built and owned live dashboards in Metabase, Redash, Looker, or equivalent — dashboards that real teams used to make real decisions.
- Has managed or mentored at least one analyst. Understands what good output looks like and can coach toward it.
- Understands statistical significance and can explain it clearly to a non-technical audience. Has killed an A/B test that was under-powered before the business acted on it.
Good to have
- Experience building event tracking schemas and working with engineering teams on instrumentation
- Familiarity with order-level and user-level data models in food delivery or quick commerce — understands cohort retention, frequency buckets, and unit economics natively
- Python or R for analysis — particularly for cohort modelling and statistical testing
- Experience setting up or owning an MMP or CRM data pipeline
- Has worked in a 0 to 1 environment where data infrastructure was being built, not inherited — is comfortable with ambiguity and missing data
Click on Apply to know more.