Inbound down across every channel? How B2B leaders can tell whether demand disappeared or moved

Inbound down across every channel? How B2B leaders can tell whether demand disappeared or moved

If inbound leads are down on your website, in paid campaigns, and through partners at the same time, you do not yet know whether buyers left the market or started choosing somewhere your dashboard does not see. On November 9, 2026, B2B SaaS marketing consultant Gaetano DiNardi posted that an unnamed B2B SaaS client's inbound leads were down 50% year over year across every channel, called the pattern a business-wide conversion and demand-capture problem, and asked who else was seeing it (original post) [1]. That 50% figure is his report about one unnamed client. It is not a verified industry statistic, not a Scale2Rev finding, and not a fact about the market as a whole.

Replies offered competing explanations: buyers shortlisting in ChatGPT or other AI tools before a site visit, branded search holding while other queries drop, stale messaging, a shrinking category, and build-versus-buy pressure from AI. Nobody in that thread gave a leader a way to tell those causes apart. Treat every one of them as a hypothesis to test. A wrong cause sends budget to the wrong fix.

Demand that disappeared versus demand that moved

Leaders who already invest in search and who use ChatGPT in a simple way still need a commercial read, not a specialist build. Disappeared means fewer buyers are in market for the category. Moved means buyers are still choosing, but the path no longer shows up as a form fill on the channels the team watches.

If the category shrank, more content on the old queries will not restore last year's lead count. If buyers now shortlist inside an AI answer and only visit to confirm a name they already trust, the site can look quiet while buying is still active. This should not start as a tracking project. Dashboards do not decide which hypothesis is live.

What changed: inbound can fall even when buyers are still in the market

When inbound leads fall on every channel at once, the instinct is to treat the market as smaller. The pattern can also mean the same buyers changed the first place they look. A marketing agency analysis of AI search argues that branded search volume can rise while unbranded volume falls, because some discovery now happens inside ChatGPT, Claude, Gemini, and Copilot and never reaches Google or Bing as a query [2]. Prospects who heard a name in a model answer often type a branded query to verify it [2]. Treat that as a hypothesis to test on your own terms, not as a law.

Non-branded queries fall, so organic and some paid discovery fall with them. Referral reports look thin if the shortlist happened off the site. Direct and branded visits can hold or rise, which is easy to miss if the team only watches total leads.

A Gartner survey of 645 B2B buyers, conducted in August and September 2025 and published on May 20, 2026, found that 45% used generative AI during a recent purchase, primarily to gather information on vendors and products [6]. That supports the idea that a shortlist can form before a visit. It does not, by itself, prove that your category has died.

Why this matters for pipeline and revenue

Pipeline forecasts that assume last year's visit-to-lead math will miss if buyers now arrive later, fewer, and more decided. Cutting content, pausing campaigns, or hiring to "make up the leads" locks in the wrong response. A disappeared category needs a harder look at the offer. A moved category needs to be named where buyers now build the list, then confirmed on the site.

A falling visit count is not, by itself, a verdict on demand. On November 10, 2026, SparkToro's Rand Fishkin published Similarweb data on visits to the world's top million websites. Those visits peaked at 2.243 trillion in the first quarter of 2022 and stood at 1.970 trillion in the third quarter of 2026, 12.2% below that peak [8]. Because about 24% more people are online, SparkToro calculated visits per web user down 29.2% [8]. The figures cover worldwide browser traffic only. They exclude apps, and the post does not prove a cause.

The useful lesson for a B2B team is more general. Judge inbound less by raw visits and more by a short scorecard: qualified actions per 100 visits, the speed of the first response, and whether buyer questions are actually answered. Those measures tell you if the visits you still get are doing commercial work.

Close rates on those visits can be higher than the volume suggests, but they are not a substitute for the diagnosis. In Octane11's Q3 2026 analysis of its B2B clients, accounts that visited through an AI engine closed a won deal at 1.22%, compared with 0.49% for organic-search accounts, 2.5 times higher, in 11 of 13 clients with enough data [4]. Siege Media, looking at 78 websites across several industries, found AI-referred visitors converted at a median 1.26 times the organic rate and matched or beat organic on 72% of those sites [7]. Its B2B SaaS subset was closer to even, at 1.05 times. Treat both as directional evidence from specific datasets, not a benchmark to match.

A diagnostic order, not a guess from a comment thread

The useful order is commercial, and it stays in this sequence. Compare branded and non-branded search trends. Check whether the company is named when buyers ask AI tools for options in the category. Check how visits turn into leads, including form friction. Review recent sales call notes for how buyers found you. Only then decide what to fix first.

Compare branded versus non-branded search trends

The branded vs non-branded search split separates people who already know the name from people still looking for a type of solution. If branded queries hold or rise while non-branded queries fall, demand may have moved into other discovery tools, with search used later to verify a name [2]. If branded and non-branded both fall, and sales hears the category mentioned less, disappearance is more plausible than a simple channel swap.

Check whether the company shows up when buyers ask AI tools for options in its category

AI search shortlisting is the second check because many buyers now research vendors before they ever open a website [6]. Ask, in ordinary language, for the kinds of options a buyer would request: tools for the job you sell, vendors for the category, alternatives to the usual names. The useful output for a leader is simple. Are you named, are you described accurately, and are you absent while competitors are listed?

Absence is a finding for that question set, not a permanent ranking. A wrong description is a different problem: buyers may visit and bounce because the answer they trusted does not match the page. Do not treat one chat session as a market study, and do not fund a content surge that repeats stale positioning.

Check conversion per visit and form friction

After you know whether discovery moved, look at the visits you still get. A lower lead count with steady completions per visit means fewer people arrived. A steady visit count with fewer completions means the page, the offer, or the form is in the way. The first is a reach problem. The second is a conversion problem. DiNardi framed the unnamed client's drop as both conversion and B2B demand capture, which is why a leader should not assume both are broken before this check [1].

Form friction is a leader-level question. Are you asking for more fields than a first conversation needs? Does the offer match someone who already shortlisted you, or does it still assume a stranger at the top of an old funnel? Do not "fix conversion" by adding popups or extra steps. If the visits you still get are later-stage, a mismatched form gives away the conversations most likely to matter.

Review recent sales call notes for how buyers found you

Search trends and AI answers still need the buyer's own account. Recent call notes, and a short question on the next calls, should record how the buyer first heard the name and what they compared you with. "ChatGPT," "a colleague," "a branded search," "we almost built it ourselves," and "we are not looking this quarter" are different diagnoses. Notes that say the category budget moved to internal builds support build-versus-buy pressure. Notes that never mention the category support shrinkage. Notes that say they already had a shortlist before the site visit support a moved path.

Only then decide what to fix first

The fifth step should follow the evidence rather than the most confident reply in a thread. If non-branded search is down, branded search is holding, you are missing from category answers, and calls say buyers arrived with a shortlist, the first fix is being found and described correctly where that shortlist is built, then making the site confirm the name. If you are named accurately, visits are steady, and completions per visit fell, fix the offer and the form, not a new channel. If branded demand, category mentions, and call volume all point to a smaller market, revise the offer and the revenue plan.

How Scale2Rev can do this work for a B2B team

Scale2Rev helps B2B companies grow inbound revenue with AI: being found in AI search such as ChatGPT, publishing demand-generating content, and converting website visitors into qualified leads. A leader does not need an in-house specialist bench to start. The partner runs the four checks, then recommends what to fix first.

The diagnosis is always done on your evidence, and the plan should follow the order your numbers support. If inbound has fallen and you need to know what to fix first, start a free trial and run that diagnosis.

Frequently asked questions

Is a drop on every channel proof that AI took the leads?

No. A shared drop means the old scoreboard is no longer enough. AI shortlisting, a weaker category, a broken form, and stale messaging can produce the same chart.

Should we pause content until the diagnosis is done?

Pause net-new campaigns that assume last year's discovery path, not the conversations already in flight. Turning off everything hides whether branded demand is still there. Keep sales talking to the leads you have while the checks run.

What if we already rank and still do not appear in AI answers?

Ranking and being named are different moments. A buyer can get a shortlist without the click your rank used to earn. The second check exists so you do not spend the quarter polishing pages nobody is asked to open.

How should we judge inbound if visits are falling?

Use a short scorecard alongside the lead count: qualified actions per 100 visits, the speed of the first response, and whether buyer questions are actually answered. SparkToro's open-web data is a reminder that fewer visits can arrive even as more people come online [8].

How long should a leader wait before changing the plan?

Long enough to see the four checks on the same date range as the lead drop, not longer than a normal forecast cycle.

References

1. Gaetano DiNardi: inbound leads down 50% YoY across every channel (Nov 9 2026). https://x.com/gaetano_nyc/status/2119958850203697436
2. How AI search shifts the branded/unbranded query split. https://techmarketing.agency/blog/ai-search-branded-vs-unbranded/
3. State of B2B AI Search Q3 2026: Traffic to Pipeline | Octane11. https://www.octane11.com/insights/insights-state-of-b2b-ai-search-q3-26/
4. Gartner Survey Finds 69% of B2B Buyers Turn to Sales Reps to Validate AI-Generated Insights (May 20, 2026). https://www.gartner.com/en/newsroom/press-releases/2026-05-20-gartner-survey-finds-sixty-nine-percent-of-b-two-b-buyers-turn-to-sales-reps-to-validate-ai-generated-insights
5. AI Traffic Conversion Rates: Are They Really 7x Better? | Siege Media. https://www.siegemedia.com/research/ai-traffic-conversion-rates
6. New Research: Traffic to the Open Web Has Fallen 12% in 5 years; Visits/web User Are Down 30%. SparkToro (Rand Fishkin, Similarweb data, Nov 10, 2026). https://sparktoro.com/blog/new-research-traffic-to-the-open-web-has-fallen-24-in-5-years/