Zero-to-One

Context

Getir was scaling fast across multiple verticals in 2020. The core business was ultra-fast grocery delivery from dark stores. Leadership saw an opportunity in bottled water, a high-frequency, subscription-like category with different unit economics and operational DNA than anything the company had built before.

I was one of eight founding members of GetirWater. I owned the business plan, franchise revenue model, growth strategy, and operations design. Hired and led a team of roughly 20.

The Problem

Water delivery looks simple until you try to make it work at scale.

Three things made this fundamentally different from Getir's core business:

Reverse logistics. Grocery delivery is one-directional: warehouse to door. Water delivery requires getting empty carboys back. Every container that doesn't return erodes unit economics. There was no infrastructure, no tracking, and no process for this.

Vehicle constraints. Carboys are heavy and fragile. The core fleet ran on two-wheelers. Water required cars only, which meant separate fleet management, different cost structures, and different routing logic.

Asset economics. A single carboy is expensive. The cost only amortizes if the container cycles through enough deliveries before end-of-life. Without visibility into where carboys were in the lifecycle (factory, with customer, in transit, lost), there was no way to manage this.

The core question: how do you build a delivery business where the thing you deliver has to come back?

Approach

Started with the economics, not the product. If carboy return rates stayed low, the vertical wouldn't be viable regardless of order volume. So the first design priority was the return loop.

QR tracking system. Designed a lifecycle tracking system: every carboy tagged at the factory, scanned at each handoff (factory → warehouse → courier → customer → return pickup → factory). This created full visibility into where assets sat and where they leaked out of the system.

The tracking data also enabled something we didn't originally plan for: identifying which customers consistently returned containers versus which ones hoarded them. That fed into delivery scheduling and deposit policy decisions.

Operations from scratch. Built the franchise model, store network design, and delivery operations in parallel. Unlike core Getir where infrastructure existed, everything here was greenfield: store locations, fleet procurement, staffing models, bundling strategy for water products.

Growth strategy. Focused on neighborhood density rather than geographic spread. Water is a repeat-purchase category. One customer ordering weekly in a dense area is worth more than five scattered first-time orders, because the delivery and return routes batch efficiently.

Results

• 45 stores operational within the first year
• 20,000+ daily orders at steady state
• 92% year-over-year order growth
• 40% increase in carboy reuse rate after QR tracking rollout

The reuse improvement alone had an outsized impact on unit economics. Each percentage point of reuse directly reduced procurement costs and waste.

What Transferred

Reverse logistics is an underappreciated capability. Most delivery businesses optimize the forward path and treat returns as an afterthought. The pattern from GetirWater, designing the return loop first and building forward operations around it, applies to any business where physical assets cycle: rental models, reusable packaging, equipment servicing. The tracking infrastructure we built wasn't technically complex. The hard part was making it the organizing principle of the operation rather than a feature bolted on later.