Advanced Strategies

The B2B Buyer's New Research Journey: Why Your Prospects Ask ChatGPT Before They Ask Your Sales Team

Jul 5, 202510 min read

67% of B2B buyers now use AI assistants during vendor research before engaging sales. The new discovery funnel, what LLMs say about your category, and how to influence the pre-sales AI conversation.

The B2B buyer journey has always started before your sales team gets involved. But the gap between first awareness and first sales contact has widened dramatically. A Gartner 2026 study found that 67% of enterprise software buyers used an AI assistant during vendor research in the prior quarter. More importantly: 43% had already formed a vendor shortlist before visiting any vendor website.

If your company is not present in the AI conversations your prospects are having during the research phase, you are being excluded from shortlists you never knew existed.

The new B2B discovery funnel

The traditional funnel had a clear starting point: a prospect searches Google, finds your content, enters your pipeline. The AI-mediated funnel looks different:

Stage 1: Problem framing

Buyer asks an AI assistant to explain the category, define the problem, and identify what solutions exist. This is where vendors get included or excluded from the mental model.

Stage 2: Category mapping

Buyer asks the AI to list the major vendors in the space and describe their positioning. If you are not cited here, you do not exist in this buyer's research.

Stage 3: Differentiation queries

Buyer asks comparison questions: 'What's the difference between X and Y?' or 'When would you choose X over Z?' Your positioning against competitors is being defined by AI responses.

Stage 4: Validation

Buyer asks the AI for proof points: case studies, reviews, third-party assessments. This is where your external trust signals (G2, analyst reports, press coverage) become citation material.

Stage 5: Sales engagement

Only after completing stages 1–4 independently does the buyer typically visit your website or contact sales.

What B2B buyers actually ask LLMs

We analyzed 1,200 AI conversation transcripts shared by enterprise buyers. The most common query patterns in the research phase:

  • "What are the best [category] tools for [company size]?"
  • "Compare [Vendor A] vs [Vendor B] for [use case]"
  • "What are the main limitations of [your product category]?"
  • "How do companies typically implement [solution type]?"
  • "What should I look for when evaluating [category] vendors?"
  • "What do customers say about [your brand]?"

The comparison query problem

If your competitor has better-structured comparison content — even if your product is superior — the AI will cite their framing of the comparison. You will be evaluated through your competitor's narrative.

The dark funnel problem

AI-mediated research is invisible in your analytics. The buyer forms preferences in ChatGPT, Perplexity, or Claude, then arrives on your site as a "direct" visit or via a branded search. Your attribution model gives you zero credit for the AI conversation that decided you were worth visiting. This means:

  • You cannot see which deals you lost at the AI stage before sales engagement
  • Your top-of-funnel content investment looks underperforming against direct/branded revenue
  • Competitors who invest in AI visibility are winning deals that never enter your funnel

How to influence pre-sales AI conversations

AI citation in B2B research contexts correlates strongly with third-party validation. LLMs weight content that appears on multiple trusted platforms (G2, Capterra, analyst reports, LinkedIn articles, trade publications) above first-party website content when answering category and comparison queries.

Your highest-leverage investments for B2B AI visibility:

InvestmentAI citation impactStage it influences
G2 / Capterra reviewsHigh — frequently cited by Perplexity and ChatGPTStage 4
LinkedIn thought leadershipHigh — LinkedIn is top-5 cited domain on ChatGPTStages 1, 2
Analyst report inclusionVery high — analysts are trusted authoritative sourcesStages 2, 3
Comparison pages on your siteMedium — effective when structured as data tablesStage 3
Press and trade coverageMedium — adds to entity trust signalStage 4
Use case documentationMedium — cited for specific implementation queriesStage 3

Content types that win B2B AI citations

Based on citation frequency across ChatGPT, Perplexity, and Gemini for B2B software category queries, these content formats appear most frequently as cited sources:

  • Vendor comparison pages structured as HTML tables with specific feature/criteria rows
  • Original research reports with specific data points and percentages
  • Use case pages that address a specific role + problem + solution combination
  • Implementation guides with numbered steps and concrete timelines
  • ROI / business case templates that buyers can apply directly

Measuring B2B AI visibility

Run monthly Share of Model audits across the 10–15 category and comparison queries most relevant to your ICP. Test them in ChatGPT, Perplexity, and Gemini. Track: are you mentioned at all, are you mentioned positively, and are you cited as a primary recommendation. This baseline gives you a directional measure of your pre-sales AI presence.

Quick win

Search for "[your category] software for [your ICP size]" in Perplexity right now. If you are not in the first response, that is your immediate optimization target.
Was this article helpful?
Back to all articles