ORCAMOBILE DSPORCA Ads · On-demand services
ORCA Ads helps food delivery, ride-hailing, grocery, and quick-commerce apps acquire users who complete that first order, then come back. Optimize toward order completion and repeat cadence, not just installs.
01 · The Growth Challenge
On-demand economics live or die on the second order. Roughly two-thirds of first-time customers never place one, and a promo-driven install that converts once is often a loss. Add local supply constraints — a customer acquired outside your delivery zone or during a courier shortage is worse than no customer — and cost per install becomes an actively misleading metric.
02 · How ORCA Supports On-Demand Growth
### Reach demand your supply can actually serve
Target down to region, city, or postal-code level so campaigns run inside your live service areas, with budget shifting by daypart to match the hours you have coverage. No spend on customers you cannot serve.
Target down to region, city, or postal-code level so campaigns run inside your live service areas, with budget shifting by daypart to match the hours you have coverage. No spend on customers you cannot serve.
Optimize toward completed orders, bookings, or revenue — and for repeat-driven models, toward the second order within 30 days rather than the first. Campaigns generally need around 50 of the chosen event per week to optimize stably.
Rank users by predicted 30-day value, then re-engage lapsed customers on inactivity windows you set. Campaigns typically recover 8–15% of a lapsed cohort within 30 days, timed to the next likely need rather than a fixed calendar.
Enter new cities and countries on evidence: most campaigns produce a reliable read on a new market within two weeks, with local bidding models rather than one market's settings copied onto another. Available across 190+ markets.
03 · Relevant Use Cases
04 · Business Outcomes
### Acquire customers who order twice
Buy for the repeat, not the promo. Advertisers moving from install-based to value-based bidding typically see 20–35% higher 30-day retention among acquired customers.
Buy for the repeat, not the promo. Advertisers moving from install-based to value-based bidding typically see 20–35% higher 30-day retention among acquired customers.
Judge budget on completed orders and revenue rather than cost per install. Teams giving ORCA Ads a revenue-linked objective commonly reach 15–30% higher ROAS within the first two months.
Bring back customers who ordered once and stalled, recovering 8–15% of a lapsed cohort within 30 days.
Get a reliable read on each new market within two weeks, so launch decisions rest on evidence rather than on a quarter of spend.
05 · Customer Proof
Roughly two-thirds of first-time customers never placed a second order, and a promo-funded first order that never repeated was a loss on the unit economics. Delivery-zone constraints made it worse — some of the acquired spend was landing on customers outside the zones the app could actually serve that week.
Targeting was narrowed to the postal codes inside live service areas, with budget shifting by daypart to match courier coverage, and the optimization event moved from the first order to the second order within 30 days.
30-day retention among newly acquired customers rose 20–35%, and lifecycle campaigns recovered 8–15% of customers who ordered once and stalled, at 40–70% less than the cost of acquiring someone new.
"The first order was never the hard part. Getting someone to reorder inside a zone we could actually staff — that's the number that changed the unit economics." — Head of Growth
Each new metro launch had been costing close to a full quarter of spend before the team knew whether local demand and supply would support the business, with no reliable read until enough bookings had accumulated to judge the market on its own.
Local bidding models were built per market rather than copying settings from an established city, the optimization event was set to a completed booking, and re-engagement campaigns were timed to each customer's typical rebooking window rather than a fixed calendar.
New markets produced a reliable read within two weeks instead of a quarter, and rebooking campaigns recovered 8–15% of customers who booked once and went quiet.
"Two weeks to know if a market works changes how many markets you're willing to try." — VP of Marketplace Growth
Next step
