What this is, who writes it, and what it refuses to claim.

Quick commerce in India is now a multi-billion-dollar market covered daily by the business press, and almost none of that coverage is based on shelf-level measurement — because the platforms don't publish it and there is no public dataset.

The gap matters because of how these platforms work. Every order is fulfilled from the dark store nearest the shopper, and each store carries its own assortment, ranks results from its own local signals and sells its own ad slots. There is no single national shelf to measure. A brand that looks healthy on a national average can be absent from a third of a city's stores, and nothing in its seller dashboard will say so, because competitor rank and rival ad placements are shopper-facing data that no seller portal contains.

We started reading individual dark stores in July 2026 because nobody else was publishing it. The site is new. The data is original.

Every figure published here is counted from the public shopper-facing shelf — the results a customer sees when they search from a given location — on a stated date, in a named city, across a stated number of dark stores.

• No seller accounts. We do not connect to, request, or receive access to any brand's seller or advertising account.
• No brand-supplied data. Nothing here is provided by a brand, a platform or an agency. No platform has reviewed or approved it.
• Nothing modelled. There are no projections, estimates or extrapolations. If we did not read it, we do not publish it.
• Per store, never averaged into a national number. Averages hide the stores where a brand is invisible, which is usually the finding that matters.

This section is longer than the previous one on purpose. Most data marketing overstates; the limits are where you should judge us.

• We do not know what anyone spends on ads. Bids are not public. We count sponsored placements owned, which is a fact, and never convert it into rupees, which would be a guess. Anyone publishing quick-commerce ad spend figures is estimating.
• We do not publish rank-spread headlines. Platforms cap how many results they return, so the lowest position we observe for a brand is a limit of what we could read — not evidence it ranks there in the full catalogue. We previously published a finding built on that number and removed it once we understood the cap. It is not coming back.
• We do not compare raw assortment size across platforms. Those same caps differ by platform, so cross-platform depth comparisons would measure the API, not the shelf. We only compare how much a platform varies between its own stores.
• A failed read is never reported as an absence. If a store cannot be read it is excluded from the sample, never recorded as “the brand is not stocked there.”
• One city on one day is a snapshot, not a trend. Every report states its date and city. Shelves change daily, and a single read cannot tell you direction.
• Category level only. No individual brand's private data, and no client's competitive set, ever appears in a public report.

The Dark Store Index is written and published by Gulshan Iyer. I run Arbhu Enterprises, an eco-friendly tableware manufacturer and exporter based in India, and work as an AI and marketing consultant. Other projects and background are at gulshaniyer.com.

I don't come from the quick-commerce industry. I'm an operator and marketer who noticed that a market everyone was writing about had no public shelf-level measurement, and who had the technical skills to go and measure it rather than wait for someone else to. Shelf X-Ray came out of that; this publication is where the category-level findings go.

I state that plainly because it tells you how to weigh what's here: judge the numbers on the method, not on my résumé. That is why the limitations section above is the longest on this page, and why a headline finding was withdrawn once it turned out to be an artefact of measurement rather than quietly left up.

There is also a commercial interest, and it is real: Shelf X-Ray is a product, and we would like brands to use it. What that interest does not get to do is change a number.

Contact: hello@darkstoreindex.com

If you are reporting on Indian quick commerce and need shelf data that isn't published yet, ask. We will run the read and share the raw numbers, including the ones that don't support a story we would prefer.

• Custom reads. Name a category, a city or a set of keywords and we will measure it.
• Raw data. The underlying per-store figures behind any published report are available on request.
• Tool access. We will set up full access with query credits for anyone reporting on this market, unconditionally — no obligation to write about us, and no obligation to mention us if you do.
• Attribution. Cite as “The Dark Store Index” with the date of the read. The date matters more than the name; these shelves change.

If a number here is wrong, we want to know, and we will correct it in place with a note saying what changed and when. Write to hello@darkstoreindex.com.

This applies to brands as well. If you believe a report misrepresents your shelf presence, tell us — we will re-read the stores and publish whatever the second read shows.