Every major report published about B2B search in 2025 told essentially the same story. AI Overviews reduced clicks, zero-click searches increased, and organic traffic declined across the industry.
That story is true, but also incomplete. Because while much of B2B lost traffic, a surprisingly large group of companies kept growing.
To understand why, Schwartz Marketing Lab (S.LAB) tracked 118 B2B SaaS companies across 10 verticals over the same 18-month window. 47% lost organic traffic. 16% stayed relatively flat. 37% grew.
The scale of the divergence was striking. 27% of companies that declined lost more than half of their organic traffic. Meanwhile, 30% of all companies in the cohort are currently more than 50% below their all-time peak.
The most common single trajectory across 118 companies? A high reached before the study window opened, followed by a decline that has not recovered.
Yet, while some companies experienced catastrophic collapse, others grew significantly.
Ramp grew 112%. Vanta grew 60%. Braze grew 60%. Several of these companies competed directly against businesses in the same categories that lost more than half their organic traffic.
The aggregate story is that AI hurt organic traffic across B2B. That part is true. But it doesn’t explain why some programs contracted while others compounded under the same conditions.
That is the question this research set out to answer.
What didn’t explain the traffic split
The obvious explanations turned out to be the wrong ones. As we worked through the data, one by one, the theories that seemed most likely stopped matching our observations.
The first hypothesis: category
The first theory was that some categories were simply hit harder than others. The data doesn’t support it.
- Security and compliance: 50% of tracked companies grew, 36% declined.
- Finance and spend management: 45% grew, 45% declined.
- HR and people operations was the hardest-hit vertical: 73% declined, 18% grew. But that 18% grew significantly.
In almost every vertical, both outcomes appeared. The category doesn’t explain which side of the split a program landed on.
The second hypothesis: domain authority or program maturity
The second hypothesis was that larger, more established programs were naturally more resilient. That didn’t hold either.
Aha! has a mature program, years of investment, strong rankings, and an extensive content library. It declined 42%.
GitHub has scale that most B2B companies would trade significant resources for. It declined 72%. Maturity didn’t protect them.
The third hypothesis: content volume
The third hypothesis was that all growing companies published a steady stream of new content. They didn’t.
The companies that continued growing weren’t publishing substantially more than the ones that declined. Publishing volume wasn’t the separating factor.
By this point, the obvious explanations had run out. Whatever separated the winners from the losers had to be something deeper than category, authority, or output.
That left one question: what did the companies that kept growing actually have in common?
Braze and Aha!: Why two similar programs produced opposite outcomes
Two companies made the answer impossible to ignore.
Braze grew organic traffic 59.9% over the 18-month window. The growth was distributed across dozens of pages, with no single page accounting for more than 15 to 18% of total gains. What stood out was that each growing page required a user to engage with it.
In other words, you cannot extract what they offer in a summary. You have to read them, think about them, and apply them to your specific situation. The program held through algorithm updates and AI Overview expansion across the full study window.
Aha! declined 42.4% during the study window. At the surface level, that might look like one more company caught by the same forces that hit everyone else. The data underneath tells a different story.
Across more than 20 individual queries, Aha!‘s pages held or improved their Google rankings while traffic on those same pages fell between 80 and 99%.
| Query | Previous position | Current position | Traffic change |
|---|---|---|---|
| product marketing | 1 | 1 | −98% |
| marketing job titles | 3 | 3 | −99% |
| user story template | 3 | 1 (improved) | −92% |
| product positioning | 2 | 2 | −95% |
| product roadmap template | 2 | 2 | −93% |
These are not marginal pages on obscure queries.
Position one: traffic fell 98%.
Position three: traffic fell 99%.
One page even improved from position three to position one. Traffic fell 92%.
Aha! is not a company that failed to invest in organic. Before the study window, it had built one of the more sophisticated content programs in the product management category. The decline wasn’t the result of abandoning SEO or producing low-quality content. The pages were good enough to rank, and they did.
The problem was that Google’s AI Overviews and other AI answer engines answered the user’s question without them having to visit the company’s website.
That creates a difficult reality for SEO teams: rankings do not always equal traffic.
Across Aha!‘s 154 template pages, net traffic loss was 60%, deeper than the 42% overall site decline. The cannibalization rate across the full program was 286%. For every unit of traffic the program created, 2.86 units were lost elsewhere in the portfolio. Each content investment was accelerating net decline rather than offsetting it.
The two programs ran in the same environment, targeted overlapping queries, and earned strong rankings. The difference was what happened after someone found them.
Braze’s content couldn’t fully satisfy the reader without a click. Aha!’s often could.
Across the cohort, we kept seeing the same pattern. The difference wasn’t category, authority, or publishing volume. It was whether the content still required someone to engage after AI could answer the question.
That realization sent us back through every growing company to determine exactly what their content had in common.
What winning companies had in common
When we compared the companies that kept growing, one pattern kept emerging.
Across five independent verticals, 61% of classifiable growing companies grew primarily through definitional and educational content built around foundational professional concepts.
Like Braze, they weren’t focused on content volume. They published content that required readers to engage. They couldn’t accomplish their goal with a summary or search result alone.
Eventually we realized we weren’t looking at better execution. We were looking at a different kind of content that didn’t disappear as AI got better at answering questions.
We started calling this durable demand. It became the first layer of a broader framework for understanding why some content programs compound while others don’t.
Every year, new practitioners enter a field and need to understand its foundational concepts for the first time. That audience replenishes continuously. Successful programs position their content to capture that demand as it arrives.
Content built around this property compounds. Content that delivered its full value in a single interaction depreciates.
The inverse is structurally fragile regardless of execution quality. A well-produced template is still a template. A thorough guide to a process that AI tools now handle on demand is still satisfiable inside the SERP. The vulnerability is structural, not algorithmic. That’s why some companies kept their rankings and still lost traffic.
The problem wasn’t visibility. It was that the user’s job had already been completed before they reached the page.
The question every B2B team should be asking
For 20 years, the central question shaping organic growth was straightforward: how do we rank higher?
That question isn’t obsolete. But it no longer covers the full problem.
The companies that continued growing through AI search weren’t just better at ranking. They built content that still required people to engage after AI could answer the initial question.
That was the first pattern. The second was how those durable assets came together to create resilient content portfolios rather than concentrating growth in a handful of fragile pages.
With that in mind, we must now ask a new question: what kind of content still requires a click after AI can provide the initial answer?
That’s the question our research kept pointing us toward. And it’s the question that increasingly separates programs that compound from programs that slowly erode.
The 2026 Organic Growth Playbook expands on these findings with:
- The five-mechanism diagnostic framework
- The Durable Demand Framework
- Deep dives into the successes and failures B2B SaaS companies
- Practical guidance for diagnosing B2B organic programs