The 1.6% Problem: What It Means That Almost No MSME Has a Working Website

One study, a striking number, and why it needs a methodology label

In December 2025, the SEO research firm SERP Forge published the results of an audit it ran across 40,000 MSME-registered businesses spanning seven Indian states and covering IT, services, e-commerce, manufacturing, and other major sectors. The businesses were randomly selected from a working set of 3.6 million MSME units registered under Udyam. The audit did not ask owners whether they had a website. It looked for one, and then tried to load it.

The results, reported by SME Street: out of 40,000 businesses, only 465, 1.6 percent, had a website listed anywhere in the record. Of those 465, only 402 actually resolved and loaded, about 1 percent of the full sample. And among those 402 working sites, only 43, roughly one in ten, scored a perfect 6 out of 6 on a basic digital-hygiene and usability checklist. The failures documented among the working sites included 5 missing HTTPS, 6 flagged "Not Secure," 38 that were not mobile-responsive, 177 with no SEO title tag, 29 with no contact information listed anywhere on the page, and 64 that were completely inaccessible despite being listed as live.

This is a single vendor study, not a peer-reviewed or government-audited count, and it should be labeled that way. SERP Forge has not published a full methodology paper describing exactly how the six hygiene checks were scored or how representative the 3.6-million-unit sampling frame is of the wider Udyam base. Treat the 1.6 percent and 1 percent figures as a contested, directional finding from one credible industry source, not a certified national statistic.

A second, larger, more established survey lands in the same neighborhood

A separate and more methodologically transparent data point exists. The Research and Information System for Developing Countries (RIS), an established Indian public-policy think tank, commissioned Kantar to field a survey of 2,882 enterprises for a study titled "MSME Digitalisation in India: Current Status and Challenges," funded by the Mastercard Center for Inclusive Growth. Its finding on web presence: 19 percent of surveyed MSMEs reported having a website or web page. The same survey found 29 percent reported using social media for marketing and 18 percent reported being engaged in e-commerce.

Nineteen percent is more than ten times SERP Forge's 1.6 percent, and the gap is worth taking seriously rather than picking a winner. The RIS figure is self-reported: an owner asked whether the business has a website may be counting an inactive domain nobody has updated in years, a marketplace storefront on IndiaMART or a similar directory that is not technically an owned site, or a Facebook business page remembered loosely as "our website." The SERP Forge figure is crawl-verified: it required an actual, listed, resolvable URL that loaded when requested. Self-report and crawl-verification are measuring adjacent but genuinely different things, a claim of digital presence versus a demonstrated one, and the size of the gap between them is itself a finding: it suggests that "having a website," as MSME owners understand and report it, is a considerably looser category than "having a website a machine can actually reach."

Whichever number is closer to the truth, both credible, independently sourced estimates land far short of a majority. Read together, not competitively, they support one conclusion with more confidence than either alone: a small minority of registered Indian MSMEs have a working, owned web presence, and depending on how strictly "working" is defined, that minority could be as low as one in a hundred or as high as one in five.

What the percentage means against the actual size of the base

Neither percentage means much without the denominator. India's Udyam Registration Portal, launched July 1, 2020, and the Udyam Assist Platform, added January 11, 2023 to formalize informal micro enterprises, together carried 4.77 crore registered MSMEs as of July 31, 2024, according to the Minister of State for MSME's written reply in the Rajya Sabha. By February 28, 2026, that count had grown to 7.83 crore, per a subsequent Ministry reply reported through the Rajya Sabha record. The Economic Survey 2024-25 separately reports that MSMEs had, as of November 26, 2024, reported employing 23.24 crore individuals, and that the Udyam Assist Platform alone had formalized over 2.39 crore previously informal micro enterprises since its January 2023 launch.

Apply either web-presence estimate to a base that size and the absolute numbers are what make the finding matter. At the RIS survey's more generous 19 percent, roughly four in five of India's formally registered enterprises, tens of millions of businesses employing a meaningful share of a workforce measured in the hundreds of millions, have no self-reported website of any kind. At SERP Forge's audited 1 to 1.6 percent, the near entirety of the registered base has no page that a crawler can actually load. This is not a story about a struggling fringe. It is a story about the working default for formal India's enterprise sector.

Being digitally active is not the same as being crawlable

A natural objection is that Indian MSMEs are not offline, they transact through platforms rather than owned websites, and government digital-commerce infrastructure has scaled fast enough that the website question may be beside the point. The Economic Survey 2024-25's own account of the Open Network for Digital Commerce (ONDC) supports the first half of that claim and undercuts the second.

As of the Survey's reporting, ONDC recorded more than 14 million transactions in a single month, November 2024, across more than 1,100 cities. Over 7 lakh sellers and service providers were active on the network across more than 600 cities and towns, with 190-plus network participants live. In the retail category alone, the platform had processed more than 4.4 crore cumulative orders across 814 districts, involving 2.83 lakh sellers. That is real, large-scale digital participation, and it is a genuine achievement of public digital infrastructure.

But transacting inside ONDC or listing on the Government e-Marketplace (GeM) is not the same as owning a page that a generative engine's crawler can independently discover, read, and cite. A seller on ONDC is discoverable inside partner buyer apps through the network's protocol, not necessarily through any URL that belongs to the seller and sits on the open web. A business can process real revenue through a government-backed digital rail while remaining, from the point of view of a system that crawls and indexes the open web the way AI answer engines do, functionally absent. Platform-mediated digital activity and first-party machine-readable presence are two different layers, and the evidence above suggests most of formal India has cleared the first without clearing the second.

Why a near-zero crawl rate specifically breaks AI answers

The distinction matters more now than it would have five years ago, because of how generative answer engines actually work. The 2024 peer-reviewed paper that introduced "Generative Engine Optimization" found that the content levers which raise a source's visibility inside a generated answer, cited statistics, quotations, and authoritative structure, are properties of a page that already exists and can be read. A business with no reachable page cannot be optimized for citation, because there is nothing for an engine to crawl, parse, or cite in the first place. It can, at best, be mentioned indirectly by a third party, a directory listing, a marketplace page, a news article, which is a weaker and far less controllable form of visibility than owning the page an engine would cite.

The stakes of that gap are rising even for businesses that clear it. A 2025 Pew Research Center browsing-panel study found that when an AI summary was present in Google search results, users clicked a traditional search result in about 8 percent of searches, against 15 percent without a summary present, and clicked a link inside the summary itself only about 1 percent of the time. Ranking and even being crawled no longer guarantees a click the way it used to; increasingly, being named inside the synthesized answer is what earns the visit, and that requires having a citable page at all.

The Indian government has, on the record, named an adjacent version of this risk. Replying to a question in the Lok Sabha in December 2025, the Minister of State for Commerce and Industry, Jitin Prasada, acknowledged that AI-powered agentic shopping, where an AI agent searches, compares, and completes a purchase with minimal human input, poses a particular challenge for retailers not integrated with major e-commerce platforms, risking reduced visibility and market access. The government's stated response leans on platform integration, through the India AI Mission and ONDC, rather than on first-party crawlable web presence. The evidence above suggests that is a distinct and, so far, largely unaddressed layer of the same problem.

What the evidence establishes, and what it doesn't

No audited, government-run, methodologically transparent national census currently measures what matters most here: what share of Udyam-registered enterprises have a working, independently reachable, minimally crawlable web page. The two available data points diverge by more than tenfold, and each has a real, specific limitation. SERP Forge's crawl audit has not published a full methodology paper, and its 3.6-million-unit sampling frame is itself a subset of a much larger and still-growing Udyam base, so how representative it is of the full 7.83 crore registered enterprises is not independently verifiable. RIS's survey is smaller, at 2,882 enterprises, and asks a self-report question that is vulnerable to the ordinary optimism and recall imprecision that surveys asking whether a business has a given asset tend to produce.

A rigorous measurement does not yet exist, and this piece does not claim to have run one. What it would look like: an independently disclosed, statistically representative sample drawn from the full Udyam Registration Portal and Udyam Assist Platform database, resolving an actual reachable domain per enterprise rather than accepting a self-report, and then separately testing whether that domain is parseable by a standard crawler and citable by a generative engine, a distinct technical question no public dataset yet answers directly. Until that study exists, the summary is directional rather than precise: two independent methods, run by different organizations for different purposes, both find that the substantial majority of India's formally registered enterprises are absent from the surface that both classic search engines and AI answer engines read. The exact percentage is contested. The direction is not.