Understanding the New SaaS Sales Funnel in the Era of AI Search

Written By

In this KiwiSaaS webinar, Whippet Digital founder Mike Morgan and technical SEO specialist Cameron Blair unpack the biggest shift they’ve seen in search and attribution in years: buyers are researching, comparing, and shortlisting inside AI tools before they ever hit your website. That change is driving declining organic clicks, fuzzier attribution, and stakeholder anxiety — unless SaaS teams update what they measure, where they build visibility, and how they regain clarity over what’s really driving pipeline.

Key themes:

  • The old, linear funnel is no longer how buyers buy
    Cameron contrasted the classic journey (keyword search → click → website → convert) with today’s reality: the journey often starts on AI platforms (ChatGPT, Perplexity, Gemini/Claude), then moves into validation (review sites like G2/Capterra, Slack communities, peer-to-peer checks), and only later reaches your website.

  • AI summaries and ‘zero-click’ behaviour are changing what traffic means
    A major theme was that AI experiences are increasingly designed to replace the click — surfacing vendor shortlists, pros/cons, pricing, and positioning directly in the interface. That creates a measurement gap: organic traffic can drop while qualified demand stays steady.

  • Shortlists are forming earlier — and you may not know you’re being excluded
    Mike emphasised how AI tools shrink large option sets into shortlists quickly, and these early shortlists heavily influence purchase outcomes. If you’re not appearing (or you’re being described inaccurately), you can lose deals before the buyer ever visits your site — and your dashboards won’t show you where you dropped out.

  • What to demote in reporting (and why)
    They recommended deprioritising several metrics that used to be ‘north star’ indicators:

    • Organic clicks: a drop may mean demand moved upstream, not that demand disappeared.

    • Last-touch attribution: it rewards the final ‘confirm’ click and ignores upstream AI research and community validation.

    • Keyword rankings: you can rank #1 in Google and still be excluded from AI recommendations.

    • Bounce rate/time on site: a short session can be a win if someone lands and immediately requests a demo.

  • What to promote instead: revenue, influence and brand signals
    They encouraged teams to shift reporting toward revenue signals (demo requests, trial sign-ups, sales-qualified leads), influence and visibility, and branded search trends. Mike also shared a telling stat from their experience: ChatGPT referral traffic converting at ~16% vs Google organic at ~1.76%, highlighting that AI-influenced visitors may arrive far more informed and high-intent.

  • Self-reported attribution: a low-tech fix for a high-tech problem
    A practical solution they pushed hard: add a ‘How did you first hear about us?’ field to demo/trial forms. Cameron noted that while tools can’t track a buyer who saw a LinkedIn post, listened to a podcast, and then asked ChatGPT before converting, the buyer will often tell you if you ask. Mike suggested improving this further by using checkboxes (not a single-choice dropdown) so buyers can indicate multiple sources used during research

  • Audit your AI visibility in 30 minutes (and track it over time)
    Mike outlined a simple self-audit:

    • Choose 10 high-intent, buying-stage queries

    • Run each across 3 platforms relevant to your market 

    • Use incognito to reduce personalisation (not eliminate it).

    • Record who shows up, how they’re described, the pros/cons, and how competitors appear.

    • Screenshot and track over time — but note: “If you’re not doing anything, it’s not going to change.”

  • Review platforms matter because they feed the machines (and inconsistency kills you)
    Cameron explained that LLMs tend to prioritise high-authority structured data, and for SaaS that often means G2 and Capterra. These platforms aren’t just for humans anymore — they’re inputs AI uses to validate and build shortlists.

  • Where to start: this week, this month, next quarter
    They closed with a practical sequencing plan:

    This week

    • Run the 10‑query AI audit across 3 platforms.

    • Add self-reported attribution (make it as granular as needed).

    • Review/update G2/Capterra profiles, especially if off-mark or >6 months old.


    This month

    • Audit top 5 high-value pages (pricing, features, comparisons) for AI-extractable structure (question-led headers + 40–60 word answer blocks).

    • Align positioning across website, LinkedIn, review platforms.

    • Identify decision-stage content gaps and prioritise plugging them.


    Next quarter

    • Build an AI visibility tracking cadence (manual or tools), focusing on meaningful signals like share of voice/topical authority.

    • Create content with proprietary input and unique data; generic content that AI can recreate without you is pointless.

    • Reset stakeholder conversations from traffic/vanity metrics to influence and pipeline quality.

Key takeaways:

  • Treat declining organic clicks as a signal to update measurement, not automatically a sign of falling demand.

  • Shift reporting focus from traffic and rankings to influence, commercial visibility, and pipeline quality.

  • Implement self-reported attribution (ideally multi-select checkboxes) to capture the hidden journey buyers will happily describe.

  • Run a recurring AI visibility audit (10 high-intent queries across 3 platforms) and track how you and competitors are framed.

  • Tighten your ‘entity’ across the ecosystem: keep website, LinkedIn, and G2/Capterra aligned, current, and review-rich — because inconsistency can mean exclusion from AI shortlists.

Watch the full discussion below to explore these insights in depth.

Next

Resources for Ops & Finance