Home/Blog/Our Organic Traffic Is Down 30% and the Pipeline Is Fine. What Happened?

Our Organic Traffic Is Down 30% and the Pipeline Is Fine. What Happened?

Organic sessions falling while total sessions and pipeline hold steady

Your traffic moved rather than vanished, and your dashboard is only counting one half of it. Organic sessions for B2B SaaS fell 29.8% this year. Total sessions fell 12.9%.

Those two figures are from the same dataset: First Page Sage's 2026 report, published 28 October, drawn from 3.7 billion sessions across 218 client sites, September 2026 snapshot.

The gap between them is the whole story, and almost every analytics setup is built to show you the first number and not the second.

What the data actually says

Measure, B2B SaaS, September 2026Value
Change in organic sessions-29.8%
Change in total sessions-12.9%
Queries returning an AI Overview82.0%
Sessions arriving from AI platforms12.1%
Zero-click share of all tracked searches75.1%

Zero-click was 54.1% in 2022. It has risen every year since and added 5.7 points in 2026 alone.

The click-through collapse is sharpest at the top. On a page carrying an AI Overview, position one fell from 27.4% to 11.6%, a relative drop of 57.7%. Position ten fell by a third. Ranking first is now worth less than half what it was, which means a site can hold every ranking it had and lose most of the traffic those rankings produced.

Why the pipeline often holds

Three things are happening at once, and they partly cancel.

Fewer visits, better visits. Somebody who clicks through after reading an AI summary has already self-qualified. They know what the category is and roughly what you do. The casual browser who would have bounced is being served on the results page and never arrives, which lowers sessions and raises conversion rate at the same time.

Research moved earlier and went invisible. Buyers now arrive having compared you already, which is the shift covered in buyers are shortlisting you inside ChatGPT. That research is real demand generation. None of it appears in your session count.

Some of the traffic genuinely came back by another door. 12.1% of sessions arriving from AI platforms is a channel that did not exist in your reporting two years ago and probably still is not broken out in it.

So the honest read of a 30% organic drop with flat pipeline is usually: demand is intact, the measurement is not.

When it is actually bad news

This is the part worth being careful about, because "the metric is wrong" is a comforting story and sometimes a false one.

The traffic drop is real demand loss, not a measurement artefact, if:

  • Pipeline created fell too, in the same period, year on year. One number, same comparison.
  • Branded search fell. AI Overviews do not usually intercept somebody searching your name. If people have stopped searching your name, that is not a counting problem.
  • Direct and referral fell with organic. A broad decline across unrelated channels is a demand signal, not a SERP change.
  • Your rankings actually dropped. Check positions before concluding anything. Lost rankings and intercepted clicks look identical in a sessions chart and need completely different responses.

If pipeline held and branded search held, you have a reporting problem. If both fell, you have a demand problem and the AI Overview story is a distraction from it.

What to change in the reporting

Add an AI-platform referral line. It is 12.1% of sessions in the benchmark and it is probably sitting inside "direct" or "other" in yours. Until it is broken out, every organic comparison you make is against an incomplete total.

Move the headline metric from sessions to identified companies. A session count measures an audience that increasingly reads without arriving. The count of named organisations researching you does not depend on a click, which is why Empiraa Signal identifies the organisation behind a visit and matches it to the account or prospect you already hold, on every plan. Twelve named companies reading your pricing page is a more useful number than four hundred sessions, and it survives the shift.

Report pipeline created beside traffic, always. The single most common error this year is a marketing review that discusses a traffic chart for twenty minutes without once checking whether pipeline moved.

The uncomfortable part

A 30% organic decline with flat pipeline is survivable and arguably fine. It is also a warning, because the mechanism that took the clicks is still strengthening: zero-click has risen every year for four years and 82% of your category's queries already carry an AI Overview.

The traffic that remains is the traffic from people who wanted to arrive. Planning on that growing again is planning on a trend reversing that has not reversed once yet.


Related reading: if three quarters of searches end without a click, what is a blog post for, a company read three pricing pages and nobody called them, and where does a B2B lead actually come from now.

Zero-click search has a definition on this site.

See how Signal identifies the companies visiting your site, included on every plan with no credits to buy.

Common questions

Why is B2B organic traffic falling in 2026?
Because answers are being delivered on the results page instead of on yours. First Page Sage, measuring 3.7 billion sessions across 218 sites, found 82% of B2B SaaS queries now return an AI Overview, and on those pages the position one click-through rate fell from 27.4% to 11.6%. The ranking did not move. The click did.
Has the traffic disappeared or moved?
Mostly moved. In the same September 2026 data, B2B SaaS organic sessions were down 29.8% while total sessions were down 12.9%, and 12.1% of sessions now arrive from AI platforms. A dashboard with an organic line and no AI-referral line reports the first number and misses the offset entirely.
Should we stop investing in SEO?
No, but measuring it by sessions is now misleading. The work that gets you cited in an AI answer is largely the work that ranked you: clear answers, specifics, structure and credibility. What changes is that a share of the return arrives as a mention rather than as a visit, and your reporting has to be able to see that.
What should replace sessions as the headline metric?
Identified companies and pipeline created, rather than people. Session counts measure an audience that increasingly reads you without arriving. The number of named organisations researching you, and what happened to them, survives the shift because it does not depend on a click.
How do we know whether our pipeline is genuinely fine?
Compare pipeline created to the same period last year rather than comparing traffic. If pipeline held while organic fell, the funnel has changed shape rather than shrunk. If both fell, the traffic drop is real demand loss and the response is different, which is why separating the two matters before anyone reacts.
Ash Brown

Ash Brown

Founder & CEO of Empiraa

Published 4 November 2026

Ready to fix the part of your business that feels messy?

Whether you're trying to execute strategy, grow pipeline, or connect the way your team works, Empiraa gives you a clearer system to run from.

GPS for strategy execution. Signal for sales growth.