Local Market Price Transparency App for Farmers and Buyers

The problem growers face when selling

Growers often sell without knowing what the same crop fetched that morning in a market twenty kilometres away, and that information gap is worth real money to whoever holds it. A middleman who knows prices are up can still offer yesterday's rate. Shoppers have the mirror problem with vegetables and fruit in city markets. Official price series exist, but they cover wholesale markets, arrive once a day at best, and are not in the units people trade in.

What already exists and the gap

• AMIS Punjab, from the Punjab Agriculture Department, publishes daily wholesale prices for over a hundred commodities from market committees across the province, on a website and an Android app, in English and Urdu.
• Zarai Mandi sends daily mandi prices over WhatsApp.
• Kissan Dost and similar agriculture apps combine market rates with advisory content.
• eNAM in India is a government trading platform that also exposes market prices.

These are top-down: official reporters, a fixed set of large markets and a daily cycle. The gap is local, bottom-up and faster: smaller markets, village-level buyers and retail prices that no official series covers, reported by the people in them, with statistics that are honest about how many reports stand behind a number.

What you build

Three roles:

• Contributor: a farmer, trader or shopper submits a price in a few taps.
• Viewer: checks today's median and range for a commodity in nearby markets, and the trend.
• Moderator: reviews flagged entries and manages the commodity and market lists.

Main flows: pick a commodity from pictures, enter price and unit, confirm market (suggested from coarse location), submit, even offline. The server aggregates per commodity, market and day, and the viewer shows the figure with its sample size.

The technically interesting part

Trust with little data. With a handful of submissions per market, one wrong entry distorts the figure visibly, and a buyer has an incentive to report low prices. Use the median rather than the mean, show the sample size next to every figure, require a minimum count before publishing, and flag submissions far from recent local values using a robust measure such as the median absolute deviation. Rate-limit per device and watch for bursts from one source. Write this up as a small statistics problem: what threshold catches planted outliers without hiding genuine price jumps on market days?

Units. A price per maund, per 5 kg bag and per crate must normalise to one comparable figure. Model units as data with conversion factors per commodity.

How to evaluate it

• Speed: time ten real users submitting a price; target under thirty seconds median.
• Accuracy: compare your daily medians against AMIS or another official series for overlapping markets; report correlation and lag. Where none exists, visit markets yourself and record ground truth.
• Robustness: inject planted bad entries into a copy of the data and report how many move the published median and how many are flagged.
• Adoption: number of active contributors and submissions per market per day during a two to four week pilot.

Suggested plan

1. Weeks 1 to 2: visit two markets, interview growers and traders, list commodities and units.
2. Weeks 3 to 5: submission flow, offline queue, commodity and market data.
3. Weeks 6 to 8: aggregation, outlier rules, trend views, Urdu interface.
4. Weeks 9 to 11: moderation queue, comparison with the official series.
5. Weeks 12 to 15: pilot with real contributors, evaluation, write-up.

Path to a product

The first users are one farmer cooperative or a group of growers linked through an agriculture extension office, plus the traders who serve them. Keep price viewing free. Revenue can come from agri-input companies and processors paying for aggregated regional price data and trends, or from an NGO or development programme funding the service for a district. Scale one district at a time; a price app with empty markets is worse than none.

Scope warning

Prices only. Not a marketplace, not payments, not logistics, not crop advice. Do not scrape official sites without permission; link to them or use published data. Design for older phones, intermittent connectivity and users who may not read English well.

Learn what you need on Sefism

• Mobile roadmap : a first cross-platform app with offline storage.
• Data analytics roadmap : medians, outliers and trend charts done properly.
• Data analytics cheat sheet : quick reference for the statistics in your evaluation.
• Firebase Documentation : offline persistence and a simple backend.
• Material Design 3 : large, clear components for low-literacy users.

Get help with this idea

This idea lives or dies on whether real people submit prices, so plan your field visits as carefully as your code. Tauseef can help you tighten the scope, choose a stack your team can finish with, and design an evaluation your examiners will respect. If you are new here, create a free account first.

• Discuss this idea with a mentor
• Send a query to Tauseef
• Book a 1:1 session