Daily markdown lists for short-dated stock in regional supermarket chains
TL;DR Daycount reads a supermarket chain's stock and sales data each night and hands every store a short list in the morning: which short-dated bread, dairy, snacks and produce to mark down, by how much, and when. Regional chains pay a monthly fee per store. It works if the discounts recover more than they give away.
Every pack of bread, curd or paneer on an Indian shelf carries a printed use-by date, and FSSAI's labelling rules require it to be declared to the day for anything with a shelf life under three months[1]. Once that date passes, the pack is a write-off. Fresh produce loses value even faster, and a 2022 national study for the food processing ministry put losses at 6–15% for fruit and 5–12% for vegetables across the chain, market stages included[2].
Inside a store the call is made by a floor manager with a sticker gun and a gut feel. Discount too late and the stock gets binned. Discount too early or too deep and you give away margin on goods that would have sold anyway. There's no public figure for how much Indian supermarkets write off, which tells you how rarely anyone bothers to measure it.
What it does, in regulatory terms: pricing and ordering software only. Daycount never handles food. Prices it suggests stay at or below the printed MRP, because selling above MRP is unlawful in India[3], and it never suggests selling anything past its use-by date.
The seed pitched real-time AI pricing and I narrowed it. Indian mid-size chains mostly use paper shelf tags and batch POS exports, and MRP rules out surge pricing on packaged goods anyway. So the useful version is a morning list and not a live price feed.
Each night, Daycount pulls stock, receipts and sales by batch where the POS records them, and estimates how many units of each short-dated line will sell at full price before their date. For the surplus it picks a markdown step (10%, 25%, 40%) and a time to apply it. Staff get the list on a phone with sticker counts per aisle. A weekly report then tells the buyer which lines keep needing markdowns, so the next order shrinks. That ordering feedback is where most of the money sits, if you ask me.
The buyer is the operations or category head of a regional supermarket chain with 10–80 stores that sells its own bakery, dairy and fresh lines. The chain pays a monthly fee per store, priced against the write-offs it removes. Large chains, the biggest with 500-plus stores[4], are more likely to build this in-house, so treat them as a later target and not your first sale.
Electronic shelf labels have gone mainstream in large Western grocers, and one vendor reports more than 650 million labels deployed with 350 clients[5]. Indian chains will meet that pressure later, and a markdown engine that works with paper stickers today can drive digital labels when they arrive. Cheap cloud forecasting has also made per-store, per-batch models affordable for a 20-store chain.
There's no public figure for Indian supermarket shrink or for spend on markdown software. The most direct signal is scale - The biggest value supermarket chain alone reported revenue of ₹59,358 Cr for FY25[4] and the largest retail conglomerate runs about 18,000 stores[6]. The post-harvest loss ranges above[2] are adjacent at best, since they cover the whole chain and not shelves, so treat them as context only.
One US vendor sells AI ordering and production planning to grocers to cut fresh shrink[7]. Another takes a different route and sells near-date stock at up to 50% off through an app at big North American grocers[8]. I found no Indian vendor focused on expiry-led markdowns, though large chains may run in-house tools. The gap is a product priced and built for regional chains on paper tags.
| Difficulty | Moderate (38/100) |
| Starting capital | ₹60–110 L |
| Time to launch | 6–9 months |
The models are not the hard part. Data is, because many chains don't record batch or expiry at goods receipt, so you may have to fix receiving before you fix anything else. Getting store staff to put stickers on in time every morning is the second problem, and it decides whether the numbers move at all.
Daycount needs no sector licence, because it's software and never handles food. Its suggestions still have to live inside two hard rules.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Don't let a model decide anything about MRP or use-by dates. Those are fixed rules in code, not predictions, and where batch data is missing the forecasts are guesses.
Want to work out the AI layer for Daycount? Talk to Yash →
I'd back a founder who has worked in retail operations or supply chain and doesn't mind sitting with store managers. If you want to sell a dashboard and walk away then this one will hurt. Prove one thing first - in five pilot stores, markdown plus waste losses on bakery and dairy fall by at least a fifth over eight weeks.
Plan for roughly ₹60–110 L and 6–9 months before launch. Run paid pilots in a handful of stores before you build anything big.
It can be, if the discounts recover more than they give away and a chain can see write-offs fall against a fair baseline. There's no public figure for Indian supermarket shrink, so your pilots have to prove it.
Yes, but expect more company. Vendors already go after fresh waste in North America. Check local pricing and date-labelling rules before you sell.
The free Daycount validation kit has the line-by-line capital breakdown, the effort scoring, the MRP and food-labelling checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for chain operators, and 5 assumptions to test. Download the kit →
Thinking about building Daycount or something close to it? The make-or-break step is getting clean batch and expiry data out of a chain's stores before you promise any savings, and that's where I can help. If you need a hand, you know where to find me. Talk to Yash →
For illustration only. Names are fictional and ideas hypothetical; sourced facts are real but may be out of date. Nothing here is legal, financial or professional advice, and Yash Khatri is not personally liable for any use of it. Read the full disclaimer.