Your saas funnel
was built for how
People Used to Buy.
Here's What Changed.
|
Summary
Latest insights. Straight to your inbox
01 – Why you should read this
The Invisible Problem
Last week, I got two LinkedIn shout-outs from people recommending me as a resource on SaaS and AI. Naturally — because I’m a marketer 😅 — I asked both of them where they’d found me.
Both said the same thing: ChatGPT had recommended me.
Not a referral. Not a search result. An AI had included me in content it generated for someone else’s newsletter and blog research. I hadn’t done anything differently. I hadn’t optimized for it. The behavior had already changed, and I was just watching it happen.
That moment sent me back to a repositioning engagement I’d done the previous year with a $10M SaaS company. We were interviewing customers — where they found the product, why they chose it over competitors. The result was striking. Nearly half of their customers said their purchase process started with an AI-powered search.
This was not a future signal. It was already the baseline.
89%
Forrester found that 89% of B2B buyers adopted generative AI within two years, naming it one of the top sources of self-guided information in every phase of their buying process.
300+
I’ve worked with 300-plus teams on positioning and growth.
The pattern I’m seeing is a funnel that’s structurally out of sync with how buyers actually behave.
~50%
We were interviewing customers of a $10M SaaS company.
Nearly half of their customers said their purchase process started with an AI-powered search.
I’ve worked with 300-plus teams on positioning and growth. The pattern I’m seeing now — across companies, stages, and categories — is a funnel that’s structurally out of sync with how buyers actually behave.
The disconect: We set up funnels so people learn about their needs and our products. But in the age of AI, people already researched their needs. They are just confirming what AI recommended is the best solution for them.
This “revolution” isn’t coming. Forrester found that 89% of B2B buyers adopted generative AI within two years, naming it one of the top sources of self-guided information in every phase of their buying process.
Most SaaS companies are treating this as a future problem. It isn’t.
In this post, I’ll cover three things:
- how to audit your funnel for AI-era buyer behavior,
- what the new acquisition, conversion, and retention logic looks like in practice,
- and the one shift to make first if you can only do one thing.
The funnel looks functional from the inside. It’s optimized for a buyer who no longer exists in quite the same form.
Work with Andrej
Is your growth stalling?
A quick discovery call to find out if positioning is the problem.
"Gen AI is revolutionizing the way business buyers approach their purchase decisions."
02 – Root cause
Pre-AI VS. POST-AI client
There’s a specific mechanism behind the softening that teams can’t explain. It’s not the product. It’s not the messaging. It’s that the funnel was designed for a buyer at a different point in their evaluation.
The old assumption: the buyer is curious. They find you through search or a paid ad, they need to understand what you do, you educate them, nurture them, and eventually convert them. The funnel is a progressive disclosure machine — slowly revealing value, building trust, nudging toward a decision.
The new reality: 6sense’s research across nearly 4,000 buyers found that LLMs are used across the middle of the buying process — comparing vendors, shortlisting, building evaluation frameworks — not just at the top. Buyers aren’t arriving at your funnel from a standing start. They’ve already been to ChatGPT or Perplexity. They’ve already shortlisted. They already know your category, your competitors, and usually your rough positioning.
They’re not curious. They’re confirming.
A funnel built to educate and nurture fails them. Not because the content is bad. Because it’s solving for the wrong job. When your demo flow still leads with “let us show you what we do,” you’re talking to a buyer who already knows what you do and came to decide whether you’re the right fit. You’re answering a question they stopped asking three steps ago.
This is the same mechanism behind the growth stalls I wrote about in ICP Drift — the funnel looks functional from the inside, but it’s been optimized for a buyer who no longer behaves the same way.
A 2026 multi-source analysis by Loganix and Averi, synthesizing 680 million AI citations and nearly 2,000 controlled sessions, found that 73% of B2B buyers now use AI tools like ChatGPT and Perplexity in their purchase research. That number is not a leading indicator. It’s current behavior.
The Granola story makes the contrast clear. Granola — an AI-powered meeting notes app — raised $125M at a $1.5B valuation and deliberately repositioned from meeting note-taker to enterprise AI platform. This wasn’t a pivot made under pressure. Granola raised from a position of strength — growing, adopted, successful in its category. But they saw where AI infrastructure was going and moved into it before the market demanded it. Most SaaS companies are doing the opposite: waiting until the old motion stops working before they ask why.
Your funnel is still optimized for how buyers discovered products in 2022.
03 – the reason
Why even Smart Teams Still Miss It
The reason teams this doesn’t get noticed isn’t negligence. It’s the lag in the metrics.
The signals that break first are the quiet ones. Close rate holds. Lead volume holds. But trial-to-paid softens a few points. Time-to-activate gets longer. Churn ticks up slightly in the 60-to-90-day cohort. None of these are alarming in isolation. Together, they’re a pattern.
The pattern is this: the buyers who used to arrive early — curious, ready to be educated — are now arriving later, already evaluated, with a specific expectation of what they’ll find. If the funnel still puts them in a nurture sequence, it loses them before they ever get the confirmation signal they came for.
Alan Zhao, co-founder of Warmly, described it from his own funnel: “AI search went from 5% to 30% of our inbound demo requests in 60 days.”
That’s not a gradual shift. That’s a structural change arriving faster than quarterly reviews catch it.
At Aimfox, we’re seeing meaningful traffic from AI referrals — AI tools consistently recommend us over most competitors in our category. What’s notable is the quality difference. The buyers who arrive via AI recommendation are already past discovery. They’ve done the comparison. They arrive expecting the product to confirm what the AI told them it would do. When the onboarding flow still starts with “here’s what LinkedIn automation is,” we’re starting three steps behind where they are.
If your ICP has shifted then the buyers arriving via AI tools may not match the profile your funnel was built for.
The mistake isn’t missing the AI shift. The mistake is assuming it only affects acquisition.
04 – The pattern
Five Signals Your Funnel Is Out of Sync
These are diagnostic signals you can check today. You don’t need new data — these are in your existing analytics.
01
Trial-to-paid conversion is flat or declining despite stable traffic.
02
Time-to-activate is getting longer, not shorter.
03
Your highest-converting traffic source is one you didn't deliberately build.
04
Your educational content is getting traffic but not converting.
05
Churn is clustering in a specific cohort.
05 – The fix
Three Systems to Rebuild Your SaaS Funnel Strategy for the AI Era
You don’t need a full funnel redesign. You need three instruments that tell you where the mismatch lives and how to fix it.
System 1: An AI visibility audit.
Check whether AI tools are recommending your product, and if so, what they say about it. Open ChatGPT, Perplexity, and Claude. Ask: “What are the best tools for [your category]?” Ask: “Compare [your product] to [top competitor].” Ask: “What do users say about [your product] for [primary use case]?”
Document what comes back. Does the AI describe your product the way you’d describe it? Does it position you against the right competitors? Does it surface the right use case for your best-fit buyer?
At Aimfox, we run this monthly. It tells us more about how our buyers are forming expectations than any survey we could run. What the AI says about your product is often what the buyer hears before they ever visit your site.
If the AI describes you wrong, every buyer it sends you arrives with a misaligned expectation. That’s not a conversion problem. It’s a positioning problem that shows up in your funnel.
System 2: A mid-funnel trust reset.
Audit your demo and trial onboarding flow against one question: does this confirm value, or does it introduce it?
AI-referred buyers arrive at confirmation intent. They’ve seen the shortlist. They chose to try your product over the others. The first 10 minutes of their onboarding experience should confirm the thing the AI told them you do well — fast, visibly, without ceremony.
At Aimfox, we restructured the entire communication and UI to surface the core outcome — a live LinkedIn automation campaign running within the first 15 minutes — rather than walking through features in sequence. Conversion from trial start to first meaningful action went up. Not because we changed the product. Because we aligned the onboarding to the intent the buyer arrived with.
For other users? We are adding an Aimfox Academy where they can learn about our webinars, knowledge base and other bonus features.
Go through your demo flow or trial onboarding and ask: at what point does the buyer get confirmation that the AI’s description was right? If the answer is “somewhere in week two,” you’ve lost the confirmation-intent buyer already.
System 3: A retention signal map.
Identify which early behaviors in your product predict renewal versus churn — and whether those signals have shifted in the last 12 months.
AI-recommended buyers often have faster time-to-value expectations. They’re not starting from scratch. They’ve been told what the product does. If it doesn’t do that thing quickly, they disengage before your onboarding sequence has a chance to work.
At Aimfox, we track three early signals that predict 90-day retention: time to first campaign launch, number of connections made in the first week, and whether the user invites a teammate within the first two sessions. These signals tell us, within 72 hours, whether a new user is heading toward retention or churn.
Map your own version of this. What does a buyer who renews do in the first week that a buyer who churns doesn’t? Compare that to the first-week behavior of buyers who arrived via AI-referred traffic versus paid or organic. If there’s a gap — and usually there is — it shows you exactly which part of the onboarding needs to change.
The funnel you built got you here. The funnel you rebuild is what gets you through what comes next.
Work with Andrej
Is your growth stalling?
A quick discovery call to find out if positioning is the problem.
An AI visibility audit.
Check what context the users get from AI and map your funnel to that context.
A mid-funnel trust reset.
AI-referred buyers arrive at confirmation intent. Time to value is more important than ever.
A retention signal map.
Identify which early behaviors in your product predict renewal versus churn — and whether those signals have shifted in the last 12 months.
06 – the ending
how the story ends
Your funnel still works. That’s the problem.
It works well enough that nobody has pulled the thread. The signals are there — softer conversion, slower activation, quiet churn in that 60-to-90-day window — but they’re each small enough to explain away.
The funnel was built for a buyer who needed to be found, educated, and convinced over several weeks. That buyer still exists. But a growing share of the buyers arriving at your funnel now are not that person. They’ve already been educated. They’ve already been convinced. They came to confirm.
The question isn’t whether your funnel needs to change.
The question is whether you’ll make that change before the softening becomes a problem that’s impossible to explain away.
