HubSpot 稱自然搜尋流量跌約 80%、營收增長，探討 AI 搜尋成效衡量

HubSpot’s reported experience is a sharp example of a problem many search teams now face: declining organic sessions do not automatically mean declining business results.

In an interview with Search Engine Journal, HubSpot CMO Kipp Bodnar said the company lost roughly 80% of its organic website traffic after Google began showing AI Overviews for more searches. Yet HubSpot continued to grow revenue. Its full-year 2025 revenue reached $3.13 billion, up 19% year over year, according to the company’s full-year results filed with the SEC.

The two figures should not be read as proof that traffic no longer matters. The traffic-loss number comes from an executive interview and is not independently verifiable through public third-party analytics. But the case usefully illustrates a broader measurement issue: when AI-generated answers satisfy informational searches before a click, raw visits can become a weaker proxy for commercial opportunity.

What HubSpot says changed

Bodnar framed the drop in Search Engine Journal’s reporting on HubSpot’s traffic loss as a disruption to the old traffic-based scoreboard, rather than evidence that the business was failing. The argument is that some lost visits came from users whose questions could be answered directly in search, while remaining visitors and other discovery paths could still produce valuable demand.

HubSpot’s reported revenue growth is independently documented in its financial results. However, public information does not establish a direct causal line from AI Overviews, AI search visibility, or any individual content initiative to that revenue growth. Many factors can affect a software company’s revenue, including product demand, sales execution, pricing, retention, and broader market conditions.

That distinction matters. This is not a universal formula in which an 80% traffic decline is harmless. It is a signal that teams should investigate whether the traffic they are losing was commercially meaningful, and whether other indicators show demand, consideration, and conversion are holding up.

From traffic volume to traffic value

A traditional SEO dashboard often puts organic sessions, rankings, and clicks at its center. Those metrics remain valuable for diagnosing visibility and technical performance. But an AI answer can influence a buyer without producing a measurable visit, particularly for early research queries.

HubSpot has described its response through its Loop Marketing framework, which includes concepts such as a Taste Profile, Buyer Sim, and answer engine optimization. Its Loop Marketing materials position the framework around adapting marketing to AI-mediated discovery and more personalized buyer journeys.

The measurement shift described by Bodnar emphasizes three areas:

• Revenue per visit: A falling session count can coincide with stronger visitor quality. Comparing revenue, qualified conversions, or pipeline against visits can reveal whether the remaining audience is more commercially valuable.
• AI share of voice: Teams can monitor whether their brand is mentioned, recommended, or cited when AI systems answer relevant prompts. This is a visibility signal, not proof that an AI mention generated revenue.
• Content-influenced pipeline: Content can contribute to later opportunities even when it is not the final tracked click. This requires connecting content exposure and engagement data with CRM or pipeline reporting where possible.

These measures do not replace search performance metrics. They add context. Rankings and sessions help explain whether content is discoverable, while pipeline and revenue measures help determine whether discovery is producing business value.

Why attribution remains difficult

AI search makes an already difficult attribution problem more visible. A person may ask an AI tool for options, see a brand recommendation, search for that brand later, and convert through a direct visit, a sales conversation, or another channel. Standard last-click reporting may credit only the final interaction.

HubSpot acknowledges this ambiguity in the reporting cited by Search Engine Journal. AI visibility signals are intended to augment conventional funnel measures, not provide a complete and certain account of causation.

The company has also promoted an 1,850% increase in qualified leads from AI-generated answers between the first quarter of 2025 and the first quarter of 2026 in its agent-first go-to-market account. That is a company-reported outcome, not a general market benchmark. Public materials do not provide a complete base size or channel mix, so the percentage should be interpreted carefully.

What search teams should do now

The practical lesson is not to abandon traffic reporting. It is to make traffic quality and non-click visibility visible alongside it.

Start by separating pages and query groups by likely intent. Informational pages that answer simple definitions or calculations may be more exposed to AI Overviews than pages helping users compare solutions, evaluate capabilities, or complete a transaction. A decline in clicks to the first group warrants a different response from a decline in high-intent product and comparison traffic.

Next, monitor a consistent set of high-value AI prompts. Record when a brand is cited or recommended, which sources appear alongside it, and whether answers accurately reflect the business. This should be a recurring measurement process, because generative answers can vary by prompt, location, product changes, and time.

Finally, connect search reporting with commercial outcomes. Review qualified leads, opportunity creation, pipeline contribution, conversion rates, and revenue by landing-page group or content theme where the available data permits. Avoid declaring causation from a correlation, but use the combined view to decide where content investment is protecting or creating demand.

If AI answers are reducing visits to your informational content, GEO Search Leads can help you assess website GEO readiness and identify optimization opportunities around discoverability and AI-era content signals. Create a GEO Search Leads account to begin building a more complete view of where your site may need attention.

Frequently Asked Questions

Did HubSpot independently prove that AI Overviews caused its 80% traffic decline?

No. The roughly 80% figure was reported by HubSpot CMO Kipp Bodnar in an interview and is based on internal company data. Public third-party analytics do not independently verify the figure or isolate AI Overviews as the sole cause.

Did HubSpot’s revenue grow despite the reported traffic decline?

Yes. HubSpot reported full-year 2025 revenue of $3.13 billion, a 19% increase year over year. That confirms revenue growth, but it does not by itself prove that AI search visibility caused the increase.

What is AI share of voice?

AI share of voice is a measure of how often and how prominently a brand appears in answers to a defined set of relevant AI prompts. It can include mentions, recommendations, and citations, depending on the measurement method.

Should marketers stop tracking organic traffic?

No. Organic traffic remains important for understanding search demand, site discovery, and content performance. It should be interpreted alongside conversion quality, pipeline, revenue, and AI visibility rather than used as the only success metric.

Conclusion

HubSpot’s account does not make traffic decline irrelevant, and its 80% figure should be treated as an unverified company claim. Still, the confirmed revenue growth alongside that account reinforces an important measurement principle: in AI-shaped search journeys, the strongest view of performance combines visits with the quality of demand, AI visibility, and downstream commercial outcomes.