The polished version of our story is on the homepage. This is the unpolished version — the problems, the near-misses, and the decisions that hurt before they helped. If you’re about to launch a fleet, read this before you do.
Challenge 1 — IoT Pairing at Scale
Pairing 100 scooters with an IoT backend sounds like a straightforward technical task. It took us two weeks longer than planned. Firmware versions mismatched with our API expectations. Network connectivity in certain zones was inconsistent. And two batches of scooters arrived with factory-locked device IDs that required direct manufacturer intervention to reset. None of this was catastrophic — but all of it delayed our revenue start date by 18 days.
What we did: We built a pre-deployment checklist — hardware testing protocol, IoT pairing verification, and a GPS accuracy check — that now runs on every vehicle before it goes live. Time to pair per unit went from 45 minutes to 12 minutes.
Challenge 2 — The Summer Heat
Nobody who hasn’t operated in the Middle East really understands what 47°C does to a fleet. Battery capacity degrades faster than manufacturer specs suggest under sustained high-temperature conditions. Rides drop sharply. Field staff work shorter shifts. Charger performance declines. In our first summer, daily rides dropped 60% from our spring peak.
What we did: We built a seasonal operations model — reduced fleet size in the hottest months, concentrated vehicles in shaded zones, adjusted pricing to stimulate off-peak morning rides, and used the downtime to focus on maintenance. We now plan for summer as a cost reduction period, not a revenue growth period.
Lesson:Your operations model must be built around your climate, not borrowed from a market with a different one. What works in Berlin does not automatically work in Doha.
Challenge 3 — Pricing We Got Wrong
Our original pricing was too low. We set it to drive adoption and ended up with high ride volume but poor revenue per ride. Riders loved us. Our unit economics didn’t. Raising prices in month 3 was uncomfortable — we lost about 15% of ride volume — but gross revenue increased by 28% because the per-ride value improved significantly.
Challenge 4 — Geofencing Too Loose
Our first service zone was drawn too generously. Scooters ended up in locations that created municipal complaints — outside permitted areas, blocking pedestrian access, parked on main roads. We received two official notices in our first month of operation. Implementing tight no-parking zones and slow zones resolved this, but it took 3 weeks of iteration to get the configuration right.
What We’d Do Differently
- Build and test IoT pairing processes before vehicles arrive
- Model seasonal demand before setting annual revenue targets
- Start with slightly higher pricing — easier to offer promotions than to raise prices
- Configure geofencing before launch day, not after the first complaint
- Hire for operational discipline first, technical skills second
