Bootstrapping a bakery on SKU-margin analytics
The Oven Vibe is a D2C bakery I built and run. Same analytical method as the lending work, applied to a business where I am also the one who has to live with the decisions.
SKU-level
margin tracking
Live
business, not a case exercise
Context
The Oven Vibe is a direct-to-consumer bakery in Sundargarh that I founded and run. I built the business and its analytics — which means every conclusion here was one I had to act on with my own money.
Problem
Small food businesses usually track revenue and rarely track margin per product. The two diverge more than owners expect: the best-selling item is frequently not the most profitable one, because ingredient cost, wastage and preparation time do not scale with popularity.
Alongside that, a physical D2C business faces a second question with real money attached — where to spend limited marketing effort. Getting that wrong is expensive in a way that is hard to see, because the counterfactual is invisible.
Approach
The same discipline as any funnel or policy problem: define the metric properly before optimising it.
Margin per SKU, computed honestly. Not revenue minus headline ingredient cost, but cost-per-unit including the components small businesses habitually ignore. That reframes the product mix question from “what sells?” to “what earns?” — and those rank differently.
Customer geography as a targeting input. Rather than spreading campaigns evenly, cluster existing customers by location to find where demand already concentrates, and target campaign effort by footfall density instead of by intuition.
What I built
- An SKU-level margin dashboard tracking sales trends, cost-per-unit, revenue contribution and margin per product — so mix decisions are made on contribution rather than on popularity.
- Customer-cluster location analysis driving footfall-density campaign targeting.
- The business itself, end to end: brand, product, operations and its live storefront.
Outcome
The bakery runs on its numbers rather than on impressions of how it is doing. Product decisions reference contribution margin; marketing spend follows demonstrated demand density.
Its real value here is as evidence: analytics is the same craft whether the subject is a ₹8–9B loan book or a bakery’s daily bake list. Define the metric properly, measure it honestly, and let it change what you do. The scale changes; the method does not.
Tools
Excel, Tableau, SQL.
Where else this applies
Small-business and early-stage unit economics anywhere: which products actually make money after true cost, and which customers are worth acquiring. The same question a marketplace asks about categories, or a SaaS company about plans.