Zero click content: how to win when nobody visits your website
TL;DR
Difficulty: Beginner
A growing share of people meet your work without visiting your website: AI summaries halve clicks from Google, and feeds show fewer people the posts that send them away. So put the whole answer where people already read, from a person rather than a company page, with no teaser link. Pair it with a newsletter, ideally one that invites readers onto your own consented list. Keep answer first pages on your site for the people who search later. Measure the outcome that needs people to remember you: branded searches in Search Console's branded queries filter, plus an open question on your enquiry form asking how they heard about you.
What to read first
Someone reads your post in full, thinks it is useful, and scrolls on. No click, no visit, nothing in analytics, and three weeks later they ring you. In this article I explain why platforms and search engines send fewer people to your site, what to publish instead (complete posts, a newsletter, answer first pages), and how to measure the result through brand search and a single question on your enquiry form, with an invented Leeds payroll bureau and its figures before and after.
The click is no longer the unit of content marketing
For fifteen years the content playbook was a chain of clicks. Write something interesting, post a teaser with a link, get the visit, capture an address, nurture, sell. Every link in that chain is weaker now. Search engines answer the question before they link, and the social platforms where business people read show fewer people the posts that try to send them away.
The search side of this is measured, and two sibling articles cover it in depth. The short version: when Pew Research Center watched the real browsing of 900 adults in the United States in March 2025, people clicked a traditional result in 8% of visits to a Google page with an AI summary, against 15% on pages without one, and clicked a link inside the summary itself in 1% of visits.1 SparkToro's analysis of a Datos clickstream panel found that 58.5% of Google searches in the United States and 59.7% in the European Union ended without any click at all.2 How answers get assembled and how a page earns a citation inside them is in beyond SEO, and the four acronyms side by side are in the complete guide to SEO, AEO, GEO and AIO. I will not repeat the mechanics here.
This article is about the other half of the problem. If a growing share of the people who meet your work never visit your website, what should you publish, where, and how do you know it worked? My answer in one line: put the whole answer where people already read, give them one channel that belongs to you, and measure the outcome that cannot happen without them remembering you, which is people searching for your name.
Worked example: the post that worked and showed nothing
Kirkgate Payroll is an invented payroll bureau in Leeds with about 250 small business clients and a website at kirkgatepayroll.example. Hannah runs it. Nobody real, every number made up. In February she wrote a careful LinkedIn post on what changes for small employers at the start of the new tax year, ended it with "full checklist on our blog, link in the comments", and watched it.
The post was seen by 1,900 people, liked by 41 and commented on by 6. The blog page received 23 visits from LinkedIn that week. Two people signed up for nothing, because there was nothing to sign up for. By the usual reading, the post failed. Three weeks later a Leeds café owner rang to ask whether Kirkgate could take over her payroll, and when asked how she had heard of them she said: "I keep seeing Hannah's posts about tax years." She had never clicked any of them.
Why platforms show fewer people your links
The incentive is easy to state. A social platform earns money while people are on it. A link is an invitation to leave. Platforms that rank posts by time spent and engagement will, almost by construction, give less reach to posts whose best part is somewhere else.
Some platforms say so openly. In November 2024 Elon Musk explained X's handling of links by saying "the algorithm recommends posts based on how much time people spend on them", noted that time spent on a link post is short, and advised people to put the link in a reply instead.3 Others deny any deliberate penalty. Adam Mosseri has said Threads does not penalise links, and LinkedIn has never published a link penalty, although it shrank the preview image on organic link posts while keeping the large one for paid promotion.3
What the measurements show is a gap, not a rule. Ordinal, a LinkedIn scheduling vendor, compared linked and unlinked posts from early 2023 to early 2026 and reports that posts with an external link reached about 26.5% fewer people at the median, with the penalty concentrated on company pages while personal profiles showed almost none.4 It calls this an observed correlation rather than a confirmed mechanism, and so do I. The old version of this article said link posts get "between 40% and 60% less reach". I cannot find a study that stands that up, so it is gone. The honest phrase is that link posts tend to reach fewer people, by an amount that varies by platform and account type, and that the person who did see the post still has to decide to leave the feed.
The second half matters more than the first. Even with no penalty at all, a reader who is comfortable in a feed does not want to open a browser tab to get the second half of a thought. Teasing the answer is a request for effort. Most people decline it.
Zero click content means the post is the product
Zero click content is content that delivers its full value where it appears. No cliffhanger, no "read the rest on my site". The complete answer, the whole checklist, the actual numbers, in the post itself.
That feels backwards to anyone trained to treat social posts as adverts for the real content. The shift is in what you think is scarce. When distribution was expensive, content was the asset and you gated it. Now distribution inside a feed costs nothing and content is plentiful. The scarce things are attention, trust and being the name someone remembers when the need arrives. Giving the answer away in full buys all three. Holding half of it back buys a few clicks and loses the rest.
Worked example: Hannah's post, before and after
Before: "The new tax year brings five changes every small employer should know about. Full checklist on our blog, link in the comments." Everything useful is one click away.
After: the post opens with the line "If you employ people in the UK, five things change for you in April." Then the five, one short paragraph each: what changes, what the employer has to do, and by when. It ends with an offer that needs no click: "If any of these is unclear for your business, message me and I will tell you which applies." No link. The same checklist also sits on the website, as a page written answer first, for the people who search for it.
The after post is the one that earns the comment from a bakery owner in Manchester asking about one of the five, a reply from Hannah, and a direct message a week later. None of that shows in website analytics. All of it is the work doing its job.
What to publish when the click does not come
Four formats carry most of the weight. Which ones you use depends on where your buyers read, not on which is fashionable.
Native posts and documents on LinkedIn
For a business that sells to other businesses in the UK, LinkedIn is usually the feed that matters. The format that works is plain: a first line that states the problem in the reader's words, short paragraphs that a phone screen can carry, the whole answer, and an ending that invites a reply rather than a click. Document posts, the swipeable PDFs, keep a framework or a step by step process on the platform in a form people save and share.
Post from a person, not only from the company page. The Ordinal data puts the measured link penalty on company pages,4 and beyond reach, people trust people. A managing director who writes plainly about the work every week is a better asset than a logo that announces things.
Newsletters: the one channel that is yours
Zero click does not mean owning nothing. A feed can change its ranking tomorrow. A list of people who asked to hear from you cannot be taken away by somebody else's algorithm. That is why the newsletter is the natural partner to zero click posts: the posts earn attention you do not own, and the newsletter turns a small part of it into a relationship you do.
There are two ways to run one, and they trade reach for ownership. A LinkedIn newsletter is open to every member, and when an edition is published "push, in app, and email notifications are sent to all your subscribers".5 The reach is easy. The subscriber list stays on LinkedIn. A newsletter sent from your own mail tool needs more effort to grow, but the list is yours, portable and independent of any platform. My usual answer is both: the LinkedIn edition for reach, and in every edition a plain invitation to the own list for the people who want more.
The own list comes with rules. Marketing mail to individuals needs their consent, with a narrow exception for existing customers who can unsubscribe in every message, under regulation 22 of the Privacy and Electronic Communications Regulations,6 enforced since 30 September 2026 by the Information Commission, formerly the ICO.7 How to design a sign up that is lawful and still converts is in consent with a plan.
Answer first pages on your own site
Your website still matters, but its job changes. It is where the people who search for your name land, and where search engines and AI assistants find a passage to quote. Pages written answer first, the question as the heading and the answer in the first sentence, serve both. The mechanics are in beyond SEO. For this article the point is simpler: the same answer you gave away in the feed belongs on your site too, so that the person who remembers you and searches later finds exactly what they half remember.
Video and communities
On short video platforms the whole value has to be in the clip, because almost nobody leaves a thirty second video to read an article. In communities and forums the rule is stricter still: answer the question in full, mention your own work only when it genuinely helps, and read each community's rules before posting. Community led growth covers that discipline in detail.
How zero click content turns into business
The path from a post to a sale rarely passes through a click on that post. It passes through memory.
The diagram runs top to bottom. Nothing in it is a click on the original post. The places you can see it are the accented step and the last one: searches for your name, direct visits and messages, and what people tell you when you ask. Those are the measures.
Measure brand search, not post clicks
When content works without a click, the cleanest outcome you can count is people searching for you by name. They cannot search for a name they do not know.
Google Search Console now separates this for you. Since November 2025 the Performance report has a branded queries filter that splits queries into branded (your name, its misspellings and closely associated products) and non branded, with the classification made by what Google calls "an internal, AI assisted system" that can misjudge some queries.8 On 11 March 2026 it became available to all eligible sites, which means top level properties with enough queries and impressions.8 Alongside it, the generative AI performance reports introduced in June 2026 show impressions in Google's AI features, though not clicks.9 Between the two you can see whether more people look for you by name, and whether your pages appear in the answers.
The number I track is a share:
Here is the clicks from branded queries in a month and the clicks from non branded queries in the same month, both from Search Console. It is a plain ratio of one part to the whole (Foundations: ratios and rates). It rises for two reasons: more people search for you, or fewer people reach you through generic searches. Read the two counts as well as the share, so you know which one moved. For changes over time the plain percentage change is enough (Foundations: percentages).
Worked example: Hannah's eight months
Hannah changed three things at the start of March. She posted twice a week from her own LinkedIn profile, complete answers with no links. She started a monthly LinkedIn newsletter with an invitation to Kirkgate's own list in every edition. And she rewrote the ten most asked questions on the website answer first. Nothing else changed, and she bought no advertising.
Across the bottom are the months, up the side the clicks from Google search, split by the branded filter. The branded line sits near 145 a month until March, then climbs to 300 by August. The non branded line drifts down by about 9% over the same period, which is the AI summary effect from the first section showing up in a small business's own data.
The branded share in January was , and in August . Branded clicks alone rose by .
Meanwhile LinkedIn sent between 20 and 35 visits a month to the website throughout, before and after. By click attribution the posting did nothing. By the measure that needs people to remember Kirkgate, it doubled.
The second instrument is the enquiry form. Hannah added one open text question, "How did you hear about us?", in March. Of the 31 enquiries from May to August, 17 named LinkedIn, Hannah's posts or the newsletter, 8 named a recommendation, 4 named Google and 2 left it blank. Most of the 17 arrived as direct visits or branded searches, which is exactly where analytics would have filed them.
Ask people how they found you
Self reported attribution is the oldest measurement in sales and the one that survives zero click best. Use an open text field, not a dropdown. A dropdown invites the first plausible option. A free answer gets "a colleague forwarded your newsletter" or "I asked ChatGPT for a payroll bureau in Leeds and it named you", which no tracking script can see.
Read the answers monthly, group them by hand into a few sources, and put the counts next to the branded search line. When both move together after a change in what you publish, you have as much evidence as a small business can get. For the wider problem of sharing that happens in private messages and never shows in analytics, see dark social.
Three further signals are worth a glance, though none deserves a dashboard. Direct messages that mention a post. The share of new enquiries who already know what you do when the first call starts. And newsletter sign ups on your own list, the one zero click number that is fully yours.
When zero click is the wrong plan
It is not right for everything. If you sell something people buy on impulse, a limited offer or a simple product with a short decision, a direct link to the product page still earns its place, and a paid campaign may do more than a month of posts.
If your business depends on search traffic to category pages, as most online shops do, zero click social does not replace the work on your site. It sits alongside it.
And if you cannot wait, it will frustrate you. Brand search is a slow measure. In the example above, nothing visible happened for eight weeks. Most businesses should run both kinds of content: some built to be complete where it appears, some built to send people somewhere specific, and an honest record of which did what.
Where I would start
Pick the one feed your buyers actually read. For most UK businesses that sell to other businesses, that is LinkedIn, from a person. Take the five questions you answer most often on calls and turn each into a complete post with no link. Start a newsletter, on LinkedIn for reach and with an invitation to your own list in every edition. Put the same answers on your website, answer first. Then switch on the branded queries filter in Search Console, add an open "how did you hear about us" question to your enquiry form, and look at both once a month for six months before you judge.
The click is not dead. It has just stopped being the place where the value of content shows up. If you would like help setting this up, I run a free workshop, and the first two weeks of work are on me.