The Most Copied Marketing Statistics on the Internet. We Traced Them to Their Sources.

You have read these numbers a hundred times. 46% of Google searches are local. 93% of online experiences begin with a search engine. 88% of people trust reviews as much as a friend.

They appear in agency pitches, in slide decks, in every “top SEO statistics” roundup, and increasingly in AI answers. So we did what nobody who copies them does. We went looking for where each one actually comes from.

Not one of the five survived the trip.

The 30 second version

• “46% of searches are local” is a remark a Google employee made at a conference in 2018. There is no published study.
• “93% of online experiences begin with a search engine” has no traceable primary source at all.
• “75% never scroll past the first page” comes from a web analytics vendor in 2010, spread by a HubSpot slide deck.
• “88% trust reviews as much as personal recommendations” is a decade old. The same survey now puts it at 42%.
• “Email returns £42 for every £1” is marketers estimating their own performance in a vendor sponsored survey from 2019.

How we traced them

For each statistic we followed the citations backwards. Most stats roundups cite another roundup, which cites another, so we kept going until we reached either an original piece of research or a dead end. Each number below gets a verdict: where it really comes from, how old it is, and whether it was ever research in the first place.

“46% of Google searches are local”

The anchor statistic of every local SEO pitch ever written. Follow it back and you do not find a study. You find a Google representative speaking at a conference called Secrets of Local Search in 2018, as the industry outlets that first reported it make clear. Google has never published the research behind it. It is a remark, repeated for eight years, hardened into a fact.

Verdict: a conference comment from 2018, not published research. Local search is plainly enormous, but nobody outside Google knows if the real number is 30% or 60%, or what it is today.

“93% of online experiences begin with a search engine”

Hundreds of pages repeat this line. We could not trace it to any primary source. The pages quoting it cite each other, or cite nothing. It is commonly attributed to research from the mid 2000s, before the iPhone existed, but no study is ever linked.

The “updated” version you sometimes see, 68%, is a BrightEdge figure from 2019. Better, but a vendor number from seven years ago, still quoted today as if someone measured it last month.

Verdict: untraceable. If a number has no findable source, it is not a statistic. It is a rumour with a percentage sign.

“75% of users never scroll past the first page”

This one has a source, and the source is the problem. It appears in a HubSpot slide deck of marketing statistics, credited to HitsLink, a web analytics service, in 2010.

2010. Before mobile first search. Before infinite scroll results. Before AI Overviews started answering the question above every result. Whatever the true number is now, it was measured in a world with none of the things that define search today.

Verdict: sixteen years old. Quoting it in 2026 is like planning around a train timetable from 2010.

“88% trust online reviews as much as personal recommendations”

This is the interesting one, because the source is real and respectable: BrightLocal’s Local Consumer Review Survey, which has run for over a decade. The high trust figures come from its mid 2010s editions, when agreement peaked at 84%.

Here is what the copies never mention. The same survey kept running. By 2020 the figure was 79%. In the current editions it has collapsed to around 42%, roughly half what it was, as fake reviews taught people scepticism.

Verdict: the source now says the opposite. Everyone kept the flattering year and stopped checking. If your pitch quotes 88%, its own source now disagrees with it.

“Email marketing returns £42 for every £1 spent”

This comes from the DMA’s Marketer Email Tracker 2019. Real organisation, real report, and worth reading how the number was made: marketers were surveyed and asked to estimate the return on their own email campaigns. The report was sponsored by an email marketing platform.

So it is not a measurement of anyone’s revenue. It is the average of what email marketers guessed about their own channel, in a study paid for by an email company, seven years ago.

Verdict: a self-estimate, not a measurement. Email is a genuinely strong channel. That is exactly why it does not need a made-up-sounding number defending it.

Why this keeps happening

Statistics pages earn links, so everyone publishes one, and the fastest way to build one is to copy the last one. The citation chain points sideways, never down. Nobody’s job depends on the number being true; it only has to sound true next to a buy button.

And it is about to get worse, because AI assistants learn from exactly these pages. Ask an AI how many searches are local and there is a fair chance you get a 2018 conference remark delivered with total confidence. The zombie numbers are becoming the answers.

What to do about it

Start: apply the click test. Before you repeat a number, or spend money because of one, try to click through to the study. Who measured it, when, and how. If the trail ends at another blog, treat the number as decoration.

Build: date every number you use. “84% trusted reviews in 2016. It is 42% now” is honest and more interesting than the zombie version. Dates are what separate research from wallpaper.

Build: prefer your own numbers. Your analytics, your enquiry rate, your review conversion. A small true number about your business beats a big stale number about everyone’s.

Scale: hold your suppliers to it. If an agency pitch leans on any statistic on this page, ask for the source. The answer tells you how the rest of the engagement will go.

The Whito view

We publish statistics pages too. The difference we can offer is that every number we publish links to something you can check, with a date on it, and when a source changes its mind, the honest move is to say so.

Marketing already asks you to take enough on faith. The numbers should not be part of that. Structure before scale, and sources before statistics.

Frequently asked questions

Are these statistics fake?

Mostly no, and that is the trap. Most began as something real: a survey, a vendor’s data, a Google employee’s remark. They became misleading through age and repetition, quoted years later without dates, methods or sources.

Do people really still trust online reviews?

Yes, but far less blindly than the copied numbers suggest. BrightLocal’s own long running survey shows agreement that reviews are trusted as much as personal recommendations falling from a peak of 84% in the mid 2010s to around 42% in recent editions.

What makes a marketing statistic worth trusting?

A named source you can click through to, a date, a described method, and ideally no financial interest in the answer. If any of those is missing, treat the number as a claim, not a fact.

Sources

• BrightLocal, Local Consumer Review Survey and historical trends in consumer review behaviour.
• DMA, Marketer Email Tracker 2019 (PDF).
• HubSpot, 120 Awesome Marketing Stats, Charts and Graphs, 2010, crediting HitsLink.
• BrightLocal, Local SEO statistics roundup, which attributes the 46% figure to a Google representative speaking in 2018.

Jacob Whitmore Founder and Lead Analyst

Jacob Whitmore founded Whito. He works on what UK businesses are actually charged for marketing, and on getting those numbers published with the date they were checked.

See Full Bio