Reading ONDC's Public Disclosure Filings: What 70+ Seller and Buyer Apps Reveal About How They Actually Rank MSME Listings

A disclosure mandate most sellers have never opened

Every seller who lists on an ONDC buyer app is, on paper, entitled to know the logic that decides whether their listing shows up first or on page four. ONDC's Network Policy, Chapter 2 Business Rules, Clause 2.3.3(k), requires every Buyer Side App to ensure search results are displayed "in a non-discriminatory manner," provided that the Buyer App "discloses, to other Network Participants and ONDC... the key parameters of the logic used for sorting/listing search results, to ensure that Sellers are able to improve their ranking on listings or results." A further proviso lets the Buyer App also publish uniform minimum standards for what gets displayed at all, so long as it applies those standards consistently.

ONDC traces this obligation to a specific external rule. In Notification No. 2023/06/06/01, dated 6 June 2023, ONDC told all Buyer Apps that Clause 5(3)(f) of the Consumer Protection (E-Commerce) Rules, 2020, requires e-commerce entities to publish "an explanation of the main parameters which, individually or collectively, are most significant in determining the ranking of goods or sellers on its platform and the relative importance of those main parameters through an easily and publicly available description drafted in plain and intelligible language." ONDC gave Buyer Apps three days: comply by 9 June 2023 or face disciplinary action, with the possibility of separate enforcement action by government authorities under the underlying rule.

This is not a soft best-practice suggestion. It is a compliance deadline tied to a central government rule, run through a network that, by ONDC's own count, had onboarded more than seven lakh sellers and service providers by 2025, the large majority of them micro, small, and medium enterprises. A rule this specific, aimed at a network this large, ought to be one of the more consequential pieces of paper an MSME seller could read before deciding which buyer app to prioritize.

What ONDC itself said the mandate does not cover

Before reading what apps actually filed, it is worth reading what ONDC said the obligation means, because the network answered this question directly. When Chapter 2 Business Rules went through public consultation, stakeholders asked ONDC to clarify the scope of Clause 2.3.3(k), including whether ranking algorithms had to be made uniform across apps and how much of the underlying logic had to be revealed. ONDC's published response, part of the official Summary of Consultation attached to the Business Rules document, is unambiguous: "ONDC also clarifies that it doesn't intend to define algorithms or actual parameters for Network Participants, and it will be left to the discretion of the Buyer Side App to decide on the algorithm to be employed." The response adds that Buyer Apps "must provide guidance to Seller Side Apps/Sellers on ways to improve searchability, without actually being required to disclose business secrets."

That distinction, between disclosing "key parameters" and disclosing an algorithm, is the central fact for a seller trying to use these filings. The rule was never designed to hand a seller a formula they could reverse-engineer. It was designed to guarantee that some named, human-readable factors exist and are published, and that the app cannot claim commercial secrecy as a reason to publish nothing at all. What counts as sufficient disclosure under that standard is left almost entirely to each Buyer App's discretion, which is exactly why the actual filings, read side by side, look so different from one another.

Reading the registry: 72 filings, wildly uneven upkeep

ONDC maintains the compliance record for this mandate in one place: the Disclosures page of its Resources and Knowledge Base, at resources.ondc.org/disclosures. It is a plain table, split into a Buyer Apps section and a Seller Apps and Logistics Service Providers section, each row linking out to the filing itself, most often a Google Doc or a page on the app's own website.

As reviewed for this piece, the Buyer Apps table lists 20 entries, from Rapidor and MyStore through Meesho, Pincode, PayTM, and Namma Yatri, with filing dates running from 7 August 2023 to 28 November 2024. The Seller Apps and Logistics Service Providers table lists 52 entries, from MyStore and uEngage through Delhivery, Shiprocket, Shadowfax, and a long tail of smaller catalogue and fulfillment platforms, filed across roughly the same window. Together that is 72 disclosure filings sitting on one government-linked registry, comfortably matching the "70-plus" figure this piece leads with, and a handful of additional rows across both tables are still marked "Pending" more than two years after the compliance deadline the network itself set.

What that table does not show is quality. A row with a green "View" link tells a seller nothing about whether the document behind it names two factors or ten, whether it was last touched in 2023, or whether the link even resolves anymore. That is a separate check, and it is the one most sellers, understandably, never run.

Three filings, three very different pictures

Pulling three filings directly from the registry shows the range.

What a seller can and cannot do with a filing like this

Read generously, the disclosures that are intact and current are genuinely useful. A seller who reads MyStore's filing learns, in plain terms, that keeping delivery SLA violations and complaint rates under 10 percent is not a soft best practice, it is the line between staying listed and being unpublished, and that F&B sellers should weight proximity and delivery speed over keyword-stuffing a title. That is an actionable, free, primary-source insight that most sellers on that app have likely never seen, because it lives on a page named "product-sorting-parameters-for-search-results" that nobody links to from the seller onboarding flow.

Read across the registry as a whole, the picture is thinner. There is no single comparable table anywhere, official or independent, that lines up what Meesho weighs against what Pincode weighs against what Magicpin weighs, because ONDC deliberately left the disclosure format to each app's discretion and never required a common template. A seller listing on three or four buyer apps simultaneously, which is the entire point of an interoperable network, cannot benchmark them against each other without opening every one of those 72 links themselves, and, per the two examples above, some of those links no longer lead anywhere useful.

The government's own MSME onboarding push has not closed that gap either. The Ministry of MSME's MSME Trade Enablement and Marketing Initiative, sanctioned with an outlay of Rs 277.35 crore over 2024 to 2027 and aimed at onboarding five lakh MSMEs onto ONDC through more than 150 awareness workshops, is built around catalogue creation, account setup, and digital literacy for sellers who are not yet on the network at all. Its published scope does not include walking existing sellers through the sorting-parameter disclosures that determine whether their catalogue, once uploaded, is ever actually seen.

What the filings show

None of this means ONDC's disclosure mandate is merely performative. The rule is real, the legal basis is real, the registry is real and centrally maintained, and at least one of the filings examined here contains a quantified rule an MSME could act on today. What the mandate has not produced is a system a typical seller can actually use: a common format, a maintained link, or any built-in path connecting the moment a seller joins ONDC to a clear account of how each app it sells on decides what a buyer sees first. A rule that requires publication but not upkeep, and that explicitly protects the underlying algorithm as a business secret, was likely always going to produce a registry this uneven.

For an MSME, the practical lesson is not to distrust ONDC's framework, it is to treat "we filed a disclosure once" as the floor, not the ceiling, of what a serious buyer app owes its sellers. The filings of more practical value are the ones that are current, specific, and written for a human to act on, rather than the ones filed in 2023 and never revisited. Verifying that costs a seller little more than twenty minutes and the willingness to open a table most people have never heard of.