- Dofollow Digest
- Posts
- Dofollow Digest #57: That 53% traffic drop? It's not what you think
Dofollow Digest #57: That 53% traffic drop? It's not what you think
Eric Carrell
February 13, 2026
Hey, it's Eric 👋
Some interesting new data on how AI is actually driving traffic to SaaS sites. Short version: it's working better than the headlines suggest.
We'll dig into it below, plus the usual roundup of what's worth reading this week.
🚨New Product: Branded Web Mentions for AI Visibility
We’re launching our new product: Branded Web Mentions - we make sure high-quality sites reference your brand in a relevant way that drives sentiment & context to your brand. These are specifically tailored to help companies get more visibility in AI searches.
🔍 DEEP DIVE: That 53% SaaS Traffic Drop? It's Not What You Think
New data from Search Engine Land analyzed 774,000+ LLM sessions hitting SaaS sites between November 2024 and December 2025. The headline number, a 53% decline from the July peak to year-end, sounds alarming.
But when you dig into the numbers, the picture gets a lot more interesting.
The decline tracks perfectly with normal B2B buying cycles. Traffic peaked in July when Q3 budgets were active and people were at their desks. It dropped through August (vacations), recovered slightly in September and October, then fell off in November and December when budgets are spent and offices empty out. Every AI platform declined in sync. ChatGPT, Copilot, Claude, all of them. That's not a channel failing. That's seasonal behavior.
Here's the part I found most interesting: while standalone AI tools saw traffic drop, workplace-embedded AI grew 20x.
Copilot went from 148 sessions in late 2024 to becoming the second-largest AI referrer to SaaS sites by year-end 2025. It captures intent where the work actually happens. Someone building a business case in Excel asks "What CRM should we use for a 20-person sales team?" and that research moment never leaves the Microsoft 365 ecosystem. ChatGPT never sees it.
The data also shows where AI sends people, and it's not where most SaaS companies want them to land. Over 41% of AI-referred traffic lands on internal search pages, not product or pricing pages. That's not because search pages are great content. It's because LLMs default to your search bar when they can't find a direct answer anywhere else on your site.
That's a crawlability problem, not an existential one.
For SaaS companies, the takeaway is clear. This isn't a story about AI discovery failing. It's about AI discovery maturing and concentrating.
The companies that will benefit are the ones making themselves findable right now: publishing pricing on crawlable pages instead of hiding it behind contact forms, creating structured comparison content that LLMs can actually parse, and treating their internal search as a real content surface rather than just a navigation tool.
We're seeing this with our own clients. The ones investing in transparent, structured, comparison-ready content are showing up when AI tools recommend solutions. The ones gating everything behind forms and publishing generic thought leadership? Invisible.
There's a link building angle here too. Blog pages with structured comparison content (think "best CRM for small teams" or "Salesforce alternatives") captured the second-largest share of AI traffic at 16.4%. These are the exact pages that benefit most from strong backlink profiles, because authority signals help determine which sources LLMs trust and cite. If your comparison content has real editorial backlinks behind it, you're far more likely to be the source an AI pulls from when a buyer asks for a recommendation.
The window to get this right is still open, but it's closing. As AI-driven discovery settles into the buying process, the brands that show up now will become the default recommendations later. That's how compounding visibility works, and it's exactly the kind of advantage worth building while competitors are busy reading doomsday headlines.
🔗 LINK ROUNDUP
Til next time,
Eric