Your Traffic Is Down but Rankings Held: The Zero-Click Playbook for B2B SaaS
You have seen the chart. GA4 shows organic sessions down again this quarter. So you open Search Console expecting to find lost rankings, and instead everything looks fine: average position is steady, impressions might even be up, and yet clicks keep sliding. Nothing is broken in the usual sense. No penalty, no deindexing, no site migration gone wrong. And now the CEO is in your DMs asking why the traffic line points down when “we still rank for everything.”
The short answer: your rankings did not fail. The click did. We published the full data on AI Overviews and B2B traffic earlier this year; this post is the operating companion - what to actually do about it, content type by content type.
Why is organic traffic down when rankings are the same?
A zero-click search is a query that ends without the user clicking through to any website: the answer is delivered on the results page itself, by an AI Overview, a featured snippet, a knowledge panel, or increasingly by an AI assistant that never touched a results page at all. Similarweb’s 2026 data puts zero-click at 68% of all Google searches, up from roughly 45% a decade ago.
The mechanism is not mysterious. Pew Research Center found that when a Google search includes an AI summary, users click a traditional search result on only 8% of searches, versus 15% when there is no summary. Links inside the AI summary itself get clicked just 1% of the time. Your page can hold position 3, earn the impression, feed the AI Overview that answers the question, and never receive the visit. That is exactly the pattern you are seeing: impressions stable or rising, position stable, clicks falling.
This is why the symptom feels so strange. Every SEO instinct says “traffic down = rankings down, go find the losing keywords.” In 2026 the more common diagnosis is: rankings held, and the results page kept the click.
Quick self-diagnosis. In Search Console, compare the last 90 days to the prior period, filtered to non-branded queries. If impressions are flat-to-up while CTR fell, you have a zero-click problem, not a rankings problem. Then segment by query intent: if the decline concentrates in how-to and definitional queries while comparison and pricing queries held, you are looking at the exact pattern this playbook addresses.
Which content is losing clicks - and which is not?
The damage is not evenly distributed, and that is the single most useful fact for planning. The Digital Bloom’s traffic crisis analysis found that 88.1% of queries triggering AI Overviews are informational, while transactional queries where users intend to purchase remain relatively protected. The same report showed the median publisher down 10% year over year in the first half of 2025, and features HubSpot - the most famous informational-content machine in SaaS - with an estimated 70-80% organic traffic decline between 2024 and 2025.
Translate that to a B2B SaaS content library:
| Content type | Typical query intent | What happens in 2026 | Your goal |
|---|---|---|---|
| How-to guides, glossaries, “what is X” | Informational | AI answers it in place; click rarely happens | Be the cited source |
| Comparisons, alternatives, “best X for Y” | Commercial | Still earns clicks; buyers verify before shortlisting | Win the click |
| Pricing, integrations, security pages | Transactional | Relatively protected; high intent | Win the click and the conversation |
| Original data, benchmarks, strong POV | Mixed | Cannot be summarized away; gets cited and shared | Be unreplaceable |
Notice what this table does not say: it does not say SEO is dead. It says informational SEO has been repriced. The how-to post that reliably delivered 2,000 sessions a month in 2023 now delivers the same answer to the same audience through an AI Overview, minus the session. The comparison page a buyer reads before shortlisting still gets its click, because nobody spends five figures on software based on a paragraph a model wrote.
The zero-click playbook, by content type
1. Informational content: switch the goal from clicks to citations
Your how-to and educational content is not wasted; it is doing its job in a place your analytics cannot see. The question changed from “does this rank?” to “does the answer engine quote us?” A page that gets cited in AI Overviews and assistant answers is still shaping which vendors enter the buyer’s consideration set - it just does not produce a session while doing it.
Operationally: keep answers extraction-friendly (a direct definitional paragraph under a question-shaped heading), keep facts specific and attributable, and track whether the engines actually cite you for your money questions. We wrote up what actually moves ChatGPT citations for a SaaS, including a prompt panel you can run monthly. Measure informational content on mentions and AI-referred traffic, not sessions. Judging it on sessions in 2026 guarantees you will kill content that is quietly working.
2. Commercial and comparison content: double down
This is where the surviving clicks live. Alternatives pages, honest comparison pages, pricing explainers, integration pages, migration guides: the buyer reading these is choosing, not learning, and transactional intent remains relatively protected from AI substitution. If your content calendar is 80% top-of-funnel how-tos and 20% commercial pages, that ratio was built for a search landscape that no longer exists. Flip real effort toward the pages that catch buyers at the shortlist moment - and make them genuinely better than the category’s lazy feature grids, because these pages now carry a larger share of your organic pipeline.
3. Build “only we can say this” content
An AI summary can replace a generic explainer because a hundred sites say the same thing. It cannot replace your churn benchmark from your own customer base, your teardown of what changed after you rewrote your onboarding, or a sharply argued position only your team would defend. Original data, named methodology, firsthand experience, and strong POV are the content classes that answer engines have to cite rather than paraphrase into anonymity - and that humans still click to read in full.
The inverse strategy is now actively punished. Google’s March 2026 core update targeted scaled content: sites hit by it lost 87% of traffic on average, with ranking drops of 60-90% for mass AI page generation without editorial review, template-with-variable-substitution at scale, and aggregators adding nothing beyond their sources. “Publish more to make up the lost clicks” is not a neutral bad idea in 2026; it is the one move that can turn a traffic decline into a collapse.
4. Fix your measurement before you fix your content
You cannot manage a channel you are measuring with the wrong instrument. Three changes:
- Track AI referrals in GA4. Sessions arriving from chatgpt.com, perplexity.ai, gemini.google.com and friends are your most direct evidence that citation-goal content works. GA4’s default grouping misses some of it; here is the full setup, including the regex for what the AI Assistant channel misses.
- Watch branded search and direct as downstream signals. A buyer who learned about you inside a ChatGPT answer does not arrive as “organic / how-to-post.” They arrive a week later searching your name or typing your URL. If branded queries and direct traffic rise while non-branded informational clicks fall, your invisible layer is working.
- Report pipeline per piece, not sessions per piece. Fewer, higher-intent visitors can produce the same or more pipeline than the old traffic firehose. If your reporting cannot connect a piece of content to the revenue behind it, every conversation about this shift becomes a vibes argument.
5. Reset stakeholder expectations - with a better chart
The hardest part of this playbook is not tactical, it is the conversation with the CEO or the board who learned to read the organic sessions chart as the health of marketing. Do not walk in with “traffic is down but it’s fine.” Walk in with a different chart: pipeline and revenue per piece, AI citations for your top 20 money questions, branded search trend, and win-rate of the commercial pages that still earn clicks. We published a section-by-section board report template for SaaS built exactly this way - records and revenue, not modeled reach. The session chart was always a proxy. Replace the proxy instead of defending it.
The one-line explanation for your CEO. “Google now answers the easy questions itself, so we get fewer visits from people who were never going to buy. The buyers who are choosing software still click, we are winning more of them, and here is the pipeline per piece to show it.” Then show the table, not the traffic chart.
Where Marqeable fits
Two pieces of this playbook are things our platform does directly. Marqeable’s revenue attribution links dollars to the exact message or piece that produced them, so “pipeline per post” is a report you pull, not a spreadsheet you assemble - which is what makes the stakeholder-expectations reset in step 5 stick. And because the visitors who still click through are the higher-intent ones, the AI chat widget answers them in seconds, captures the lead, and books the meeting instead of letting your most valuable remaining traffic bounce off a static page. We’re in private beta with a small early cohort - get early access.
Frequently asked questions
Why is my organic traffic down when my rankings are the same?
Because ranking no longer guarantees a click. Pew Research Center found users click a traditional result on only 8% of searches with an AI summary present, versus 15% without one, and links inside the summary get clicked about 1% of the time. Your page holds its position and earns the impression, but the results page delivers the answer.
Is SEO dead for B2B SaaS in 2026?
No, and this post deliberately avoids that framing. Informational SEO has been repriced: the click it used to earn often never happens, so its job is now citation and influence. Commercial and comparison SEO still earns clicks and still converts. The work shifted from one goal to two.
What content still earns clicks?
Comparison pages, alternatives pages, pricing, integrations, security and migration content - anything read at the shortlist moment. Transactional queries remain relatively protected from AI Overviews, and buyers verify vendors before spending money.
How do I measure content in a zero-click world?
Four instruments: AI citations for your money questions, AI-referred sessions in GA4, branded search and direct traffic as downstream demand, and pipeline per piece. Sessions alone will tell you working content is failing.
Should I publish more content to recover the traffic?
Not by scaling volume. The March 2026 core update cost sites using mass AI generation without editorial review 87% of their traffic on average. Publish fewer, denser pieces with original data and firsthand experience - content the engines must cite and humans still choose to read.
The bottom line
The traffic chart broke because the search landscape split in two. Informational queries increasingly end on the results page: 68% of Google searches are now zero-click, and when an AI Overview appears only 8% of searchers click a traditional result. Commercial queries still send clicks, and they were always the ones that paid the bills. So the playbook is not “fight for the old chart back.” It is: point informational content at citations, concentrate click-effort on commercial intent, build what only you can say, instrument AI referrals and branded demand, and report pipeline per piece to the people who used to read the sessions graph. Rankings held. Now make the metrics catch up to how buyers actually find you.
Marqeable runs your campaigns, answers every visitor, text, and email in seconds, and turns them into booked jobs and meetings - even at 9pm on a Saturday. We’re in private beta with a small early cohort. Get early access
