Why Every "No Website" Statistic You've Read Is Outdated

Search how many small businesses have no website and one page of results will give you four different answers, none of which agrees with the others and most of which trace back to the same small set of surveys.

We measure this directly rather than asking. Across 285,554 businesses in 22 countries, 31.1% have no website — and the more useful finding is how little a single global figure means once you look underneath it.

They are self-reported. Most come from surveys asking business owners whether they have a website. Owners answer yes when they have a Facebook page, a directory listing, or a site that stopped working two years ago. Every one of those inflates the answer in the same direction.

They are American. The most-quoted numbers describe US small businesses and get applied worldwide. Our measurements put the United States at 18.1% and India at 41.2% — same measurement, entirely different market.

They are old, and rarely dated. Figures written in one year get quoted in the next without the original date attached, so a statistic can be five years old by the time it reaches you looking current.

They average across categories that have nothing in common. A single national figure covers dentists at 7.6% and hardware stores at 50.2%. The average describes neither, and the average is what gets quoted.

This is the substantive point rather than a complaint about sourcing. The variation within any market dwarfs the difference between markets.

All from businesses with a verified website check. Category variation is the largest of the four.

Category is the widest axis, which means a national statistic is the least useful number available to anyone actually prospecting. Knowing your country's rate tells you nothing about whether to work barbers or dentists this month, and that is the only decision the number could have informed.

The method is unglamorous: read the website field on a business's own listing, one business at a time, and record what is there. No survey, no sample weighting, no estimate.

It produces figures that are conservative rather than flattering, and in a specific direction. A business whose listing points at a Facebook page, a Justdial profile or a domain that no longer loads counts as having a website. Those businesses have no site in any sense that matters to an agency, and every rate we publish counts them as served.

It also produces figures that change. When a later check finds a business has built a site, the record is corrected — which makes our own headline number smaller over time. That is the correct behaviour for a measurement and the wrong behaviour for a marketing statistic, which is part of why the circulating numbers do not move.

Worth knowing, because the two main sources of error do not cancel out — they point in opposite directions and neither is small.

Survey figures understate the gap. An owner with a Facebook page or a site that died in 2023 answers yes when asked whether they have a website, because from their point of view they do. So a self-reported number reports fewer businesses without a site than actually lack a usable one.

Measured figures like ours also understate it, for the mirror reason: we count anything in the listing's website field as a website, including that same Facebook page and that same dead domain.

So both methods err in the same direction, and the true first-website opportunity — businesses with nothing an agency would recognise as a website — is larger than every published figure including ours. What differs is the size of the error, and a measured field is at least checkable one business at a time.

The one error that runs the other way is age. A figure from several years ago overstates today's gap, because businesses have been building sites throughout. An undated survey combines both errors and there is no way to tell what is left.

None of the national ones, including ours.

The number that should drive a decision is the count of businesses in your category, in your city, with no website today. Not a percentage and not a national figure — a count, because a count tells you whether there is a quarter of work there and a percentage does not.

Thirty-three restaurants with no website in Jaipur is a decision. "27% of small businesses lack a website" is a sentence for a pitch deck. The second one has been quoted for years precisely because it is too vague to be checked.