Four years ago, ShareDeal was a sketch on a napkin: what if neighbors pooled orders to get the same prices wholesalers do? In rural and peri-urban Bangladesh, the gap between farmgate and consumer prices is brutal. We thought tech could compress it.
1. The product is a wedge, not the whole knife
We didn't try to be Amazon. The first version of ShareDeal did one thing: a Flutter app where a single "leader" places a bulk order, friends and neighbors join in, and we deliver. That focus mattered. Every time we tried to widen the wedge — fancier search, social feed, recommendations — adoption flatlined.
2. Capital follows traction, traction follows a metric
Raising BDT 70M in seed looked like a milestone from outside, but inside the team it was a side effect. We obsessed over one number: active group leaders per week. Once that curve turned consistently up, the conversation with investors changed completely.
3. Build a system, not a hero
As CTO, my job stopped being "write the code" and started being "make sure the code can be written by 15 people without conflict." That meant ruthless conventions: GetX everywhere, one folder structure, one way to call the API, one CI pipeline. Boring is a feature when you're trying to ship five releases a month.
What I'd do differently
- Invest in observability earlier. We flew blind for the first year and paid for it in production fires.
- Hire one product person before five engineers. The bottleneck is rarely throughput.
- Write the customer support script before launch — it's the best product spec you'll ever read.
The hardest part of a startup isn't building the product. It's building the discipline to keep it small while the world keeps asking you to make it bigger.
ShareDeal is now at 40K+ installs, growing across districts I'd never visited before this journey started. The napkin is long gone, but the original wedge — pool the order, share the deal — is still the only thing on the home screen that matters.