The Parking App
GPS-powered parking. Dhaka had questions.
We wanted to solve Dhaka's traffic crisis by letting drivers find empty parking spots via GPS in real time. Bold idea. The parking spots didn't get the memo.
THE PROBLEM WE CHASED
Dhaka is one of the most congested cities in the world. A significant portion of that congestion comes from drivers cruising for parking — circling blocks, double-parking, jamming intersections. We believed a GPS-based real-time parking finder could meaningfully cut that wasted time and reduce friction for urban drivers.
HOW IT UNFOLDED
Early ideation
The Spark
Anyone who's driven in Dhaka knows the drill: you circle the same block three times, eventually park somewhere illegal, and accept the chaos as a fixed cost of city life. We thought: GPS exists. Real-time data exists. Why can't someone just show me where the empty spots are?
Product design
The Concept
The idea was clean. Drivers open the app, see a live map of available parking spots nearby — commercial lots, street-side slots, private spaces opened up by owners. Spot owners earn a little. Drivers save time and fuel. The city breathes slightly better. Simple value prop, obvious problem.
Research & field testing
The Reality Check
The moment we went from whiteboard to street, the gap became undeniable. Real-time GPS accuracy in dense urban areas is rough. Dhaka's parking 'infrastructure' is mostly informal — spots don't have fixed coordinates. Owners were skeptical about listing spaces. And crucially, the 'available spot' data would need constant human verification to be trustworthy. We didn't have a scalable way to do that.
Trying to salvage it
The Pivot Attempts
We tried scoping down to just commercial lots with fixed GPS tags. Then to a pre-booking model for verified paid lots. Each pivot narrowed the TAM and increased the coordination burden. We were essentially trying to build a marketplace with no liquidity on either side and no clear path to bootstrapping supply.
Decision to stop
The Shutdown
We never shipped a public product. We stopped before burning more time on a problem that needed city-level infrastructure changes, not just a better app. Sometimes the best product decision is knowing when the prerequisite conditions don't exist yet.
WHAT I TOOK FROM IT
Infrastructure problems need infrastructure solutions
A great app cannot fix a fundamentally unstructured physical system. Dhaka's parking situation needs policy, signage, designated zones — not GPS pings. We were building a layer-3 solution on top of a layer-0 problem.
Supply-side marketplace cold-starts are brutal
Even if demand was there, convincing individual parking space owners to participate, tag their spot, and keep data updated is a massive activation problem. We underestimated this by a lot.
Early field validation beats product polish
We spent too long on the concept and not enough time standing on Dhaka streets watching where people actually parked and why. Physical-world products need physical-world research from day one.
Timing matters — but so does ecosystem readiness
Unlike Rongbuzz (where we were early to a market that eventually arrived), with the parking app the market may never arrive in the form we imagined. Dhaka may skip GPS-parking entirely and go straight to designated ride-drop zones.
“Sometimes the best product decision is recognizing when the prerequisite conditions for your product don't exist yet — and knowing that no amount of clever UX can fix that.”