A major software company just demonstrated something that should unsettle anyone who treats website traffic as a proxy for business health: traffic can collapse and revenue can still go up.
HubSpot, the customer relationship management and marketing platform used by thousands of small businesses, reported a traffic decline of roughly 80% on parts of its site. At the same time, the company's marketing leadership says revenue tied to that traffic did not fall in proportion — in some cases it held steady or grew. The explanation offered was not a mystery algorithm trick. It was a shift in where buyers actually discover and research products before they ever land on a company's website.
What happened
HubSpot built much of its marketing engine over the past decade on search engine optimization — blog posts, guides, and glossary pages designed to rank on Google and pull in visitors who would eventually become customers. That playbook worked for years and was widely copied across the software industry. It is one reason HubSpot's own blog became a case study taught in marketing courses.
The traffic drop tracks with a broader change in how search works. Google's AI Overviews, ChatGPT, Perplexity, and other AI tools increasingly answer user questions directly, without sending a click to the website that originally supplied the information. A user searching "what is a CRM" may now get a full answer inside the search results or a chat window and never visit HubSpot.com at all. The content still gets used — by an AI model — but the visit, and the ad impressions or lead-capture forms that used to come with it, disappears.
HubSpot's response, according to its marketing leadership, was to stop treating total site visits as the north star metric and instead track revenue generated per visit, along with how many leads or customers arrive through AI referral links specifically. The company also says it has leaned harder into channels it controls directly — email lists, communities, and in-person or branded events — rather than depending on search engines to deliver an audience each month.
Why it matters
This fits a pattern showing up across the content and SEO industry over the past year. Multiple publishers and software companies have reported double-digit traffic declines coinciding with the rollout of AI-generated search summaries. The common thread is that informational content — the "what is X" and "how do I do Y" articles that used to reliably pull in visitors — is the category hit hardest, because that's exactly the kind of question AI tools are built to answer directly.
What's different in HubSpot's case is the revenue data. Most companies reporting traffic declines have stayed quiet about whether revenue followed traffic downward. HubSpot's willingness to say revenue didn't collapse suggests that not all traffic was equally valuable to begin with — a lot of SEO traffic may have been browsers, researchers, and students, not buyers.
What this means for small businesses
If your business has invested in a content marketing or blog strategy aimed at driving SEO traffic, the metric you've been reporting to yourself — monthly visitors — may no longer tell you much about business health. A more useful set of numbers includes leads generated per visit, which channels those leads came from, and whether AI tools like ChatGPT or Perplexity are sending any referral traffic at all (most analytics platforms can now segment this).
This also raises the stakes for owning direct relationships with customers — email lists, text subscribers, loyalty programs — rather than renting attention from Google each month. A search engine's algorithm change can zero out years of SEO work overnight; a customer list you built does not evaporate the same way.
What to watch
Watch whether other mid-size SaaS and e-commerce companies start publishing similar traffic-versus-revenue breakdowns in earnings calls or investor updates over the next two quarters. Also track Google's and OpenAI's stated plans around citing and linking to source websites within AI answers — any changes there will directly affect how much of this lost traffic is recoverable.
Bottom line
Traffic numbers alone no longer reliably indicate whether a marketing strategy is working; businesses that track revenue and leads by source, rather than total visits, will have a clearer picture of what AI-driven search changes are actually costing or saving them.