Amazon Grocery Offers September 2026: 1 in 40 Dry Fruit Prices Cut

We read ten Amazon India search shelves on 29 September 2026 - 296 listings - and counted something simpler than prices: how many tiles carried a discount badge at all. On the dry fruits shelf, one listing in forty did. On the webcam and treadmill shelves, every single listing did.

That is not a small difference in degree. It means the answer to "is there an offer on this?" is largely settled before you search, by which aisle you are standing in.

Every figure above was read on the same day, from the same host, logged out, with ASINs deduped. The shelves with 15 or 16 listings returned that many on the first page of results; the rest were capped at 40. A note on one row: the "tissue box" search returns mostly tissue holders - car visor napkin holders, wooden box holders - rather than packs of paper, so read that line as the holder shelf.

The dry fruits shelf is the extreme. Of 40 listings, exactly one showed a discount: a Rs. 349 dates-and-pista bite pack against a stated MRP of Rs. 400, a 13% badge, with 3,895 ratings. The other 39 tiles showed a price and nothing struck through.

The prices were not high - a median of Rs. 354, with the discounted listing sitting at almost exactly that median. This is a shelf that sells at its price. Waiting for a sale on it, or hunting the tiles for a badge, is wasted effort in a way that the same behaviour on the treadmill shelf is not.

The second pattern is counter-intuitive. On the green tea shelf only 8 of 40 listings carried a badge - but the median badge among those eight was 67%. On toilet cleaner, 11 of 40 carried one and the median was 70%. Compare that with the vacuum flask shelf, where 39 of 40 tiles carried a badge and the median was 25%.

So the shelves that almost never advertise a discount advertise a very large one when they do, and the shelves that always advertise one mostly advertise a modest one. A 70% badge on a Rs. 136 toilet cleaner comes from a stated MRP of Rs. 2,720 on a Rs. 136 bottle; a 25% badge on a flask comes from an MRP a few hundred rupees above the price. The two numbers are doing completely different jobs.

This is why we price things per unit rather than by badge. On the garbage bag shelf we read the same morning, the third-cheapest bag in the read carried a 1% badge and the single most expensive bag carried a 42% one - the full working is in our garbage bag price-per-bag read.

An absent badge is frequently an absent MRP rather than a discount of zero. 18 of the 40 tissue tiles printed no struck-through figure whatsoever, and so did 10 of the 40 garbage bag tiles. One dry fruits tile and one toilet cleaner tile did the same.

A seller is free to list without stating an MRP, and many do. The practical effect is that on those shelves the tiles are not comparable on "discount" in any sense - there is no second number to compare against - and the only comparison left is price, or better, price per unit of the thing.

Two green tea listings from one brand displayed a 100% badge: Rs. 261 against a stated MRP of Rs. 174,000, and Rs. 240 against Rs. 160,000. Those MRPs are roughly 666 times the selling prices.

This is the display fault we documented separately: Amazon's price block sometimes carries a per-unit rate or a figure recorded in paise rather than a genuine M.R.P., and the badge is then computed from it. It is a data artefact, not an offer. We excluded both listings from every median on this page. The full explanation, with eleven tiles verified one at a time, is in what that second figure on an Amazon tile actually means.

It is tempting to summarise this as things-you-rebuy do not go on sale and things-you-buy-once do. Four shelves fit that neatly - dry fruits, green tea, toilet cleaner and air freshener at 2% to 40% incidence, against webcams, treadmills, projectors and flasks at 93% to 100%.

The garbage bag shelf breaks it. Bin bags are about as consumable as a product gets, and 29 of 40 tiles carried a badge. So the honest statement is narrower: badge incidence varies from 2% to 100% depending on the shelf, it cannot be predicted from the price of the goods, and it is worth checking once for a category you buy regularly rather than assuming either way.

• On a low-incidence shelf, buy when you need it. A shelf where 1 in 40 tiles is discounted is not going to reward waiting, and the median price of the discounted listing was the same as the shelf median anyway.
• On a high-incidence shelf, the badge tells you nothing on its own. When 39 of 40 listings carry one, carrying one is not a distinction - compare the prices, not the percentages.
• Treat a very large badge on a cheap consumable as a question, not an answer. A 70% or 95% badge on a Rs. 130 bottle comes from a stated MRP many times the price, and stating a high MRP is lawful and commonplace.
• Divide by the unit. Bags, tea bags, millilitres, grams - the per-unit rate is the only figure on these shelves that is comparable between two tiles.