Glovo Georgia

What came in

Couriers were leaving for the competitor, and every other number followed: no-shows, churn, the cost of hiring again and again. The brief was a vague courier-experience mandate with an internal NPS. We took on the product and operational metrics of the courier side, reporting to the head of operations, with program budgets agreed directly with the general manager and the payout economics in our hands: budgets, savings, reconciliation.

What we did

• Started with research: more than thirty focus groups across courier segments, to learn what actually hurt and why people left.
• Rebuilt competitor analysis from a small manual sample into a market-intelligence system: a live view of what couriers earn on the other platform and how it moves with rain, load and time of day, at 95% confidence, within five percent. Pay rates were adjusted at the right moments to win the couriers who worked on both.
• Designed and launched Glovo Plus in Georgia, one of the first three pilot markets. The original concept was flat cash benefits. We added real ones, from free internet and fuel to leasing, but the couriers' oldest complaint was not money: slots are booked a week ahead and an order cannot be declined without a penalty. So the top of a four-tier program got the right to decline orders and to book next week's slots an hour earlier, behind a threshold set high.
• Built churn forecasting: eight metrics per courier from five pipelines into one health rating, with a reactivation flow that fires before a courier is gone.
• Built the dashboards the operation ran on: product, marketing, operations and finance, including a heat map of the competitor's orders across the city against our own.

What changed

Within three months the two upper tiers were taking every available slot, because their orders per hour had risen so much. The same volume ran on about 3,000 couriers instead of 5,000, and Georgia went about ten months without hiring a single courier, with the marketing and operating costs of hiring saved with it. No-shows −80%. Retention +35% at one month and +50% at three, churn −30%, NPS +45%. Orders per courier-hour from 1.7 to 3.2, delivery time −20%. Georgia was the only successful launch of the three pilots; the program's structure and the market-intelligence software then scaled to other countries.

What was built

• Tier-based loyalty program with partner rewards
• Payout economics and reconciliation
• Market-intelligence system
• Churn forecasting and reactivation flow
• Automated courier communications, onboarding and reactivation
• Product, operations and finance dashboards