What actually works now
Marketing in 2026, with the source for every claim
One hundred and forty three pages on how people find and choose a business now, what the platforms actually publish about it, and which of the industry's favourite numbers survive being looked up. Free, in full, no email.
What this is.
Research carried out between 28 July and 8 August 2026, published as eight parts. Every claim is checked against a primary source before it goes on a page. Sources are labelled first party, meaning the company's, regulator's or researcher's own publication, or second party, meaning a vendor study or a press report.
No percentage appears without what it is a percentage of. Every claim carries the date of its source. Where two credible sources disagree, both are printed and the disagreement is left standing rather than resolved. Where a page could not be retrieved, that is said on the page rather than filled with a secondary source.
It ends with eight questions nobody has published an answer to, which is the most useful page in it.
68.01%
Of US Google searches ended with no click in early 2026, up from 60.45 percent in 2024. The UK is 69.5 percent
Part one
70 to 56
The fall in UK adults exploring new websites, measured by the regulator, in a single year
Part one
2 in 3
Of the widely repeated platform claims we checked did not survive contact with the primary source
Part seven
55 to 60%
Of all recorded email opens are affected by Apple's proxy, and are not reliable opens
Part three
15 Sep
The date a Cloudflare setting many owners switched on starts blocking Googlebot
Part four
The eight parts.
Each part stands on its own. Nothing assumes you have read what came before it.
- 01
What changed
Feeds stopped being lists of the people you follow, and the platforms have published less about that than you would think. Where the claim that a third of people start searching on TikTok actually comes from, and the regulator-measured findings that make a better argument than it does.
- 02
The platforms
Eight chapters, one per platform, each built from what the platform itself publishes rather than from what the industry repeats about it. Instagram, TikTok, Facebook, YouTube, LinkedIn, X, Reddit and communities, and messaging and dark social.
- 03
Email
The channel where you own the list, whose headline metric stopped meaning anything in 2021 and whose rules became conditions of delivery in 2024. What Google, Yahoo and Microsoft now require, and why the open rate measures a proxy server.
- 04
Search and AI
The questions owners actually type, with the answer in the first two lines and the evidence underneath. How often Google really shows an AI answer, how AI answers pick who to quote, whether llms.txt does anything, how local ranking works, and what happens to Googlebot on 15 September 2026.
- 05
Paid and traditional
No ad platform publishes what its advertising costs, so every benchmark in circulation is somebody's client book. What the largest controlled experiment ever run on paid search found, and the only channel with a published, non-negotiable price.
- 06
Measurement
Six platforms use the word impression for six different events. Your analytics has a channel for AI traffic and will not publish what is in it. And the single best-evidenced link between a marketing number and revenue turns out to be about ratings.
- 07
Myths and facts
The four ways a marketing statistic goes wrong, the full register of thirty one claims checked against the source, what survived, and the eight questions nobody has published an answer to.
- 08
The toolkit
Twenty-two checks traceable to a primary source, an order of work with no assumptions about how much time you have, eleven things to stop doing, a one page monthly scorecard, and six checks that take five minutes each.
How to check anything in it.
Four questions will resolve almost any marketing statistic. Percent of what. Measured when. By whom, and what do they sell. And where is that number in their document.
The last one catches more than the other three combined, because most claims fail the moment somebody opens the source. Almost nobody opens the source. That is the entire advantage available here, and it is free.