Why Does Nobody Read the B2B Content You Publish?

Why does nobody read the B2B content you publish?

Because publishing stopped being distribution and nobody told the content team. For fifteen years search delivered the audience for free, so the job ended at the publish button. That arrangement has weakened measurably, and most B2B content operations have not changed a single step in response.

We run our own publishing operation alongside building sites for clients, so this is not a theory we picked up from a conference. It is the problem we have to solve on our own blog every week.

Here is what actually changed, and what has to happen after publishing now.

What does the data actually show?

A measurable drop in clicks when an AI summary is present. Pew Research Center, in a study published on 22 July 2025, tracked the browsing of 900 US adults through March 2025 and analysed 68,879 unique Google searches, of which 12,593 produced an AI summary.

The finding is stark. Users clicked a traditional search result in 8 percent of visits when an AI summary appeared, against 15 percent when one did not, which is close to half. They clicked the AI summary's own cited links in just 1 percent of visits.

The session behaviour is the part that should worry a content team more. Pew found people ended their browsing session entirely on 26 percent of pages with an AI summary, against 16 percent of pages with only traditional results. The search satisfied them and the journey stopped.

How widespread is this already?

Widespread enough to plan around. Pew found that around one in five Google searches in March 2025 produced an AI summary, which it puts at 18 percent. That is not every query, and it is not a small minority either.

The shape matters as much as the share. Pew reports that the median AI summary was 67 words, ranging from 7 to 369, and that 88 percent of summaries cited three or more sources. So the summary is short, it names several sources, and almost nobody clicks any of them.

Being cited is therefore worth something for visibility and very little for traffic. Those used to be the same thing. They are not anymore, and a content plan that measures only sessions will read this change as a failure of its writing.

So is SEO finished?

No, and anyone selling you that line is selling something. Four in five searches still produce no AI summary, and the pages that get cited in the ones that do are overwhelmingly pages that already rank. Search remains the largest single channel for most B2B companies.

What has changed is that search is no longer sufficient on its own. It used to be a distribution channel that also happened to require good content. It is now a visibility channel that requires separate distribution work to turn into anything.

The honest summary is that the free ride ended. That is uncomfortable rather than catastrophic. Our comparison of GEO versus SEO covers the strategic side of the same shift.

What does a real distribution plan look like?

A named list of places and people, written before the article is published, with a person responsible for each. Not a checklist of channels. A list of specific communities, specific newsletters, specific individuals who would find this genuinely useful.

For a B2B company that usually means four or five entries. The industry Slack or Discord where your buyers actually talk. Two or three newsletters in your space. The five people whose repost would matter. Your own email list. That is the whole plan and it fits on a card.

What makes it work is writing it before publication, because the plan should shape the piece. If the only place you can honestly send it is your own newsletter, that tells you something about the piece while there is still time to change it.

Why do most distribution attempts fail?

Because they are broadcasts pretending to be participation. Dropping a link into a community you have never contributed to is not distribution, it is an advert with worse targeting, and everyone in that community can tell.

The version that works is slower and it requires a person rather than a schedule. Someone from your company who is actually in that community, posting the substance of the idea rather than the link, and treating the article as a footnote for people who want more.

That is genuinely more expensive than a Buffer queue. It is also the only version we have seen produce anything. Our piece on founder led content covers who should be doing this and why it usually cannot be delegated to an agency.

Does your own list matter more now?

Considerably more, because it is the only channel whose reach you control. Every other route to your audience runs through a platform that can change its mind. An email list is the closest thing to owning the distribution, and it is the most neglected asset in most B2B marketing.

The requirement is that the list be worth being on. A list built by gating a whitepaper is a list of people who wanted a whitepaper, and mailing them your blog is a way to shrink it. A list built by people choosing to hear from you is a distribution channel.

This is slow work with no shortcut. It is also the only part of your distribution that will still be there after the next platform change. Gating a whitepaper to grow the list faster is the trade that quietly ruins it.

Should you publish less?

Almost certainly, if distribution is the constraint. If you can meaningfully distribute two pieces a month and you are publishing eight, six of them are going nowhere and consuming the effort that would have made the other two land.

The arithmetic is not complicated and teams resist it anyway, because volume is visible and distribution is not. A content calendar with eight rows looks like productivity. Two pieces that were actually read look like underachievement until you check what happened.

We publish a lot ourselves, and the reason is specific to running a studio blog where breadth of coverage is part of the point. That is not the right model for most B2B companies and we would not recommend copying it.

What should you do about being cited rather than clicked?

Treat the citation as the outcome and design the sentence for it. If a summary will quote one line from your piece, make sure there is a line worth quoting that names you. Specific numbers, named frameworks and clear definitions get quoted. Adjectives do not.

This is a real shift in how to write a paragraph. The old goal was to make someone click. The new secondary goal is to be the sentence that appears when nobody clicks, because brand exposure inside an answer still does something even when traffic does not follow.

Do not over rotate on it, though. Pew's data shows most searches still have no summary at all. Write for readers first and make the quotable sentence a bonus rather than the structure.

How do you measure any of this?

By counting conversations rather than sessions. Did anyone reply, share it internally, mention it on a sales call, or reference it in an inbound enquiry. Those are weak signals individually and they are the ones that correlate with anything commercial.

Keep tracking traffic, because a collapse tells you something. Just stop treating it as the scoreboard. A piece read by 200 of the right people and quoted in three buying conversations beat a piece with 4,000 sessions from the wrong query, and no analytics dashboard will tell you that.

Add a simple question to your sales call notes: where did you first come across us. It is the cheapest attribution available and it is more honest than most models. Our piece on B2B SaaS attribution covers the limits of the sophisticated alternatives.

What would we change first?

Halve the publishing volume and put the recovered time into a written distribution plan per piece, with one named owner. Do that for a quarter before changing anything else.

Most teams discover in week three that they do not have anywhere to distribute to, which is the actual finding. Building that presence is the work, and it takes months, and there is no version of this that starts next Monday and pays off next Friday.

If you want help thinking through what your content operation should look like now that publishing is not distribution, we are happy to walk through it. Find us at phoenix.studio.