State of the Buyer.

The State of the Buyer.

AI is the new front door.

The single biggest shift between 2024 and mid-2026 is that the buyer's first action stopped being a Google search. In B2B software, 51% of buyers now begin vendor research inside an AI chatbot, up from 29% twelve months earlier (Averi B2B SaaS citation benchmarks, 2026). Across all B2B buyers, 73% report using AI tools like ChatGPT, Perplexity, or Claude in their purchase research (Averi / PR Newswire syndication, 2026). In consumer behavior, roughly 45% of online shoppers engaged an AI assistant during their most recent purchase journey, up from 18% in 2024 (Digital Applied, 2026).

The Big Four are accelerating this. Deloitte launched Zora AI on NVIDIA's stack as a multi-agent platform across finance, procurement, sales, and marketing. EY launched EY.ai Agentic for Sales in March 2026 with Snowflake and Canva, and a $1B Microsoft partnership followed in May. KPMG put governed agents inside Clara, now in use by more than 95,000 auditors. McKinsey's Lilli processes 500,000+ prompts a month and is in active use by 70% of staff.

The product is not the news. The downstream effect is. The largest professional services firms in the world have replaced the first few hours of analyst work with agents. Their clients have noticed and are demanding the same internally. When the buyer's research analyst is an agent, the buyer's research surface is whatever sources that agent considers credible.

Attention scarcity is structural, not generational.

Microsoft Research's 2026 panel of 28,000 screen users across 19 countries put sustained on-screen focus at 43 seconds per session, with users switching applications, tabs, or devices an average of 1,847 times per day (Microsoft Research, 2026). The 71% of viewers who decide whether to stay or leave inside the first three seconds of a piece of content are not Gen Z. They are everyone, and they are reading on phones in line at coffee shops or scrolling between calls.

The Goldfish stat that gets quoted at every B2B conference (the "8-second attention span") is not real. What is real is selective attention. People will still read a 4,000-word piece if it answers the question they actually have. They will close a 600-word landing page in three seconds if the first line is generic.

AEO has reorganized the research phase.

Answer Engine Optimization, the practice of structuring content so AI models cite it as an answer, has gone from fringe SEO discipline to required practice in 18 months. The reason is mechanical. Google's AI Overviews now appear on more than 25% of all searches, up from 13% twelve months earlier, and they cut downstream organic CTR by 38% on triggered queries (Search Engine Journal, 2026). Google's AI Mode pushes zero-click rates to 93% at 100M users.

For brands, two facts matter. First, citation patterns across platforms barely overlap: only 11% of domains are cited by both ChatGPT and Perplexity. ChatGPT favors Wikipedia and encyclopedic structure. Perplexity favors Reddit and forum-style content. Google AI Overviews favor YouTube and multi-modal. Second, brands are 6.5x more likely to be cited via third-party sources than their own domain. The path to AI citation runs through earned media, third-party benchmarks, and category authority, not the company blog.

Trust collapse forced a verification economy.

The 2026 Edelman Trust Barometer found 70% of respondents unwilling or hesitant to trust someone with different values, facts, or problem-solving approaches. That number is for people. Brands are in worse shape. 82% of US consumers reported encountering fake reviews in the past year, 93% double-check AI-generated recommendations before purchasing, and 27% always verify (Omnisend / Digital Commerce 360, April 2026). Amazon removed 275 million fake reviews in 2024 alone.

The behavioral response is concrete. Buyers cross-reference. They look for "proof of presence" (real video of a real business, behind-the-scenes content, founder-attached posts). They trust influencers and peers more than institutions. 62% of consumers who trust a food or lifestyle influencer say they would consider trusting a brand they currently distrust if that influencer vouched for it. Endorsement networks now do work that brand websites used to do.

Stakeholders tripled in five years.

The buying group is the single most-changed structural element of the B2B world. Forrester's 2026 buyer survey put the typical business buying decision at 13 internal stakeholders plus 9 external influencers (Forrester, 2026). For complex purchases, those numbers grow. Gartner's earlier benchmark of "6 to 10" looks light compared to the current data (Gartner B2B buying journey). 74% of buying teams report unhealthy conflict during the decision process (Gartner Sales Survey, May 2025).

In B2C, the committee shows up as the household. Parents check Reddit while a teenager checks TikTok. The decision still belongs to the buyer, but the influence pool is wider, and it includes algorithms.

Channel fragmentation killed the funnel diagram.

The McKinsey B2B Pulse counted 10.2 channels in the average buying journey, up from 5 in 2016 (McKinsey, B2B Pulse). LinkedIn newsletter engagement is up 47% year-over-year, with more than 500,000 active subscribers across the platform. Podcast ad spend is projected to hit $5.5B in 2026. TikTok and YouTube Shorts are top discovery surfaces for product research under 35, and increasingly above 35 too.

The implication is not that you need to be everywhere. The implication is that attribution models built on last-click have stopped working. 70% to 80% of the B2B buyer journey is now untracked (Similarweb dark funnel analysis, 2026). Marketing teams that still report on "marketing-sourced pipeline" by last-touch are reporting on the 20% they can see.

The new path: "I think I have a problem" to "I picked a vendor."

The classic six-stage funnel (awareness, interest, consideration, evaluation, purchase, loyalty) still describes the buyer's internal state, but the external behavior at each stage has changed. Today the path looks closer to this.

A buyer notices a problem. They open ChatGPT or Perplexity and ask a question framed as their problem, not as a product category. The AI returns three to five brand candidates and a synthesis of the trade-offs. The buyer cross-checks two of those brands on Reddit, LinkedIn, and YouTube, looking for real users and real failure modes. They watch a founder's post or a podcast clip. They tell a peer in a private Slack what they're looking at. Eventually they hit one or two vendor websites. By the time they fill out a contact form, they have a shortlist of two, a budget, and an internal champion.

The form fill, the demo request, the "first touch," all of it happens in the last 20% of the journey.

The B2B Bridge.

The buying committee is 8 to 12, sometimes much larger.

The headline number from Forrester is 13 internal stakeholders and 9 external influencers per decision. The practical median, across mid-market deals our team sees and across published research, is 8 to 12. Gartner's recent data on stakeholder count tracks at 6 to 10 for complex solutions, and 6.8 average per deal. Five years ago the same number was 4.6. The buying group nearly doubled.

Worth highlighting: 94% of buyers in groups of six or more report clear benefits from the size of the group. Buyers like the group. Sellers do not. 74% of those same groups report unhealthy conflict in the decision process. The conflict is structural. Each stakeholder brings four to five pieces of independent research, often pulled from different AI tools that cite different sources. The vendor's job is no longer "convince the buyer." It is "give the buying group a shared reference document they can agree on."

Sales cycles are 25% longer than three years ago.

Average B2B SaaS cycle: 134 days, up from 107 (Ziellab benchmark, 2026). Mid-market median: 84 days. True enterprise: 6 to 18 months. Deals over $250K ACV often run 180 to 365 days. Healthcare deals with new customers are at four months minimum, and almost half push past seven (Brand Auditors, 2026).

Three forces explain the elongation. First, the bigger buying group needs more rounds of internal alignment. Second, security and compliance have become table stakes at small deal sizes. A $30K deal at a mid-sized buyer routinely triggers a security questionnaire that takes two to four weeks. Third, finance is back. After two years of "growth at all costs" budgets, CFOs are reviewing every deal above $25K. Most contracts now run through a 30 to 45 day legal cycle with redlines.

The takeaway for marketing: if your nurture sequence is built around a 60-day cycle, you are spamming the buyer for the second half of their journey. The cycle is twice that.

"Rep-free" is the majority preference.

61% of B2B buyers now prefer a rep-free buying experience overall (Gartner Sales Survey, June 2025). 60% have used a paid trial, sandbox, or usage-based pilot to evaluate a vendor before any human contact. 83% of the buying journey now happens away from sales reps (Demand Gen Report, 2026).

The corollary: "request a demo" is no longer a top-of-funnel CTA. It is a bottom-of-funnel CTA buried inside an 80-day research process. Most B2B sites still treat it like the only goal. The pages that work in 2026 give the buyer something to do without a rep: a product tour, a sandbox, a calculator, a benchmark, a comparison page, a real pricing range. If a buyer leaves your site without knowing whether you're in their budget, they assume you are out of it.

AI does the first cut of vendor research.

89% of B2B buyers use generative AI for self-guided research (Forrester, 2026). 51% start in an AI chatbot for B2B software specifically. AI search traffic converts at 14.2% vs Google organic's 2.8%, a 5.1x advantage. Claude users convert highest at 16.8%.

The catch: only 22% of marketers currently track AI visibility, and only 25.7% plan to develop content specifically for AI citation. The gap is the opportunity. The brands that get cited by ChatGPT, Perplexity, and Google AI Mode for their core category questions in the next 12 months will own the new front door. The brands that do not will spend H2 2027 buying ads against their own category.

The dark funnel is the funnel.

Forrester and Gartner both peg 70% to 80% of the B2B buyer journey as happening in untracked channels: private Slack communities, peer DMs, AI chats, podcasts, newsletters, YouTube. The dark funnel has four primary surfaces in 2026: private communities, AI research tools, peer networks, and dark social (link sharing via email, text, DM).

For most B2B companies, this means three things. The traffic and pipeline you measure are a 20% slice of the actual buying activity around your category. The marketing investments with the lowest measurable ROI on a last-touch model (LinkedIn personal posts, podcasts, community sponsorship, newsletters) are often doing the most work. And the "marketing-sourced pipeline" number you report to the board is structurally undercounting your wins by 3x to 5x.

What trust signals work, and which ones do not.

Some signals move pipeline in 2026.

• Case studies with named companies, named problems, named numbers. Case studies rank as the most effective trust content in the consideration and decision phases, with 78% effectiveness, and companies that publish regularly generate 45% more qualified leads.
• Public roadmaps. A real roadmap (not "vision deck" roadmap) signals that you ship and you do not hide.
• Founder-led content. Personal LinkedIn profiles generate 8x the engagement of company pages. LinkedIn users are 3x more likely to trust an individual than a brand (Workflows.io, 2026). Founder-led B2B companies outperform non-founder-led counterparts by 2.1x in total shareholder returns.
• Specific numbers in posts. "We grew from $0 to $400K ARR in 9 months" beats "we drove explosive growth." The buyer is filtering for proof of presence and proof of operator.
• Third-party citations on AI-friendly sources. HBR, the Information, industry trade press, IDC, Gartner, vendor-agnostic benchmarks. Brands are cited 6.5x more from third-party sources than their own domain.

Some signals look like they should work, and do not.

• Testimonial walls of one-line quotes. 92% of buyers are more likely to purchase after reading a trusted review, but generic "great team" testimonials are skimmed past. The trust signal is specificity, not volume.
• Generic G2 score badges. Buyers acknowledge G2's utility as a comparison map but express substantial skepticism about review authenticity, often cross-referencing TrustRadius and Reddit before trusting the number.
• Logo soups. "These 200 brands trust us" with no context. The buyer reads it as "these 200 brands signed something at some point." Three named case studies beat 200 logos.
• Awards and certifications without dates. A 2019 "Best of" award read in 2026 is a negative signal.

Vertical cuts.

SaaS. Cycles 84 to 134 days mid-market, 6 to 18 months enterprise. Buyer is product-led-trial-trained. Self-service evaluation is required. AI search penetration is highest here: 51% start in a chatbot. Win pattern: free-tier or sandbox, public roadmap, founder content, AI-cited benchmark or category authority piece.

Manufacturing and industrial. Slower digital shift, but the shift is real. 39% of B2B buyers now spend $500K+ per order via self-service ecommerce or remote interactions, up from 28% two years ago. 20% will place $1M+ orders digitally (McKinsey B2B Pulse). Win pattern: configurable quotes online, technical spec PDFs that AI can cite, distributor enablement.

Professional services. Buyer expects partner-level expertise during the evaluation, not after the engagement starts. AI agents from Deloitte (Zora), EY (EY.ai Agentic), KPMG (Clara), and McKinsey (Lilli) have reset the floor. If a mid-market firm cannot produce an "agentic" first-cut analysis for free during the pitch, the buyer assumes the firm is behind. Win pattern: visible operator content, named partner bios, public POVs on your category.

Healthcare. 12 to 15 stakeholders typical. Cycles 4 to 7+ months. Clinical "E-E-A-T" (Experience, Expertise, Authoritativeness, Trust) is the most valuable currency, and short-form clinician-to-clinician video is outperforming corporate webinars. LinkedIn, email, webinars, SEO content. Self-service sandboxes increasingly expected.

The B2C Bridge.

Mobile is not a channel, it is the substrate.

Mobile accounts for roughly 57% to 60% of global ecommerce sales in 2026 (Redstag, 2026). It drives 75% of traffic. During the 2025 holiday season, US mobile share hit 56.4% of online sales, up from 54.5% in 2024, and crossed 60% on Thanksgiving for the first time. Spending in 2026 is projected between $2.51T and $4T globally, depending on methodology.

The gap matters: mobile drives 75% of traffic but only 57% of sales. Desktop converts 1.5x to 2x higher than mobile web. Brands that solve the mobile conversion problem (faster pages, fewer form fields, native pay, simplified product detail) capture disproportionate share. Brands that treat mobile as a smaller copy of desktop continue to lose value at the cart.

AI shopping assistants are now a real channel.

Roughly 45% of online shoppers used an AI assistant during their most recent purchase journey, up from 18% in 2024. ChatGPT signed shopping integrations with Target, Instacart, and DoorDash in late 2025 and early 2026. Perplexity launched its shopping agent in November 2024 with Shopify integration (Shopify, 2026). McKinsey forecasts $900B to $1T in US retail revenue from agentic commerce by 2030, $3T to $5T globally.

Note the false start: OpenAI abandoned its in-chat Instant Checkout in March 2026 after low adoption, with fewer than a dozen Shopify merchants ever integrating it (Modern Retail, 2026). Amazon sued Perplexity for unauthorized purchases. The agentic commerce future is real, but the rails are unsettled. Optimize content and product feeds for AI assistants to discover, recommend, and link, but do not assume in-chat checkout in H2 2026 yet.

Ad surfaces are shifting under the SERP.

Google AI Overviews now appear on 25%+ of all searches, doubling from 13% twelve months earlier. Zero-click search rose from 54% to 72%. Organic CTR drops 61% when an AI Overview appears. The combined effect is that the click economy is moving up the page into AI summaries, and ad surfaces are moving inside chat. Chat-native ads (sponsored answers, sponsored product cards in Perplexity, ChatGPT promoted retailers) are an emerging category and likely to formalize through H2 2026 and 2027.

For B2C brands, the practical move is to treat structured data, product schema, and merchant feeds as critical AI infrastructure, not SEO afterthoughts. The buyer is asking ChatGPT or Perplexity "what running shoe should I get for plantar fasciitis under $150" and the model is pulling structured data, reviews, and editorial citations to compose an answer. If your PDP does not have clean schema and a citable review surface, you are absent from the answer.

Trust signal evolution.

The fake review crisis is concrete. 82% of US consumers encountered a fake review in the past 12 months. 93% double-check AI-generated recommendations before buying. 27% always verify. 30% of online reviews on average are fake or ungenuine, and fake reviews are growing 12.1% faster than total reviews. Trustpilot pulled 4.5M reviews in 2024. Google blocked 240M. Amazon blocked 275M.

The buyer's defensive behavior is now standard:

• Reverse-search products and brand names to find Reddit discussion threads.
• Check the brand's TikTok or Instagram for evidence of a real warehouse, real team, real customers.
• Look for the founder's face attached to a real LinkedIn or Twitter account.
• Watch for video proof of product in use, not just lifestyle imagery.

The brands that win consumer trust in 2026 lean into "proof of presence." Real warehouse footage. Founder LinkedIn posts. Behind-the-scenes content. Customer-shot UGC with named handles. Specific numbers ("we shipped 41,000 orders last quarter from our Pennsylvania facility") instead of vague claims.

Industry cuts.

DTC. Customer loyalty is now the most-tracked metric. Loyal customers convert at 60% to 70%, vs 5% to 20% for new prospects (Emarsys / Yotpo, 2026). Successful brands are shifting budget from top-of-funnel acquisition to retention and community. "Channel agnosticism" is the operating model: DTC is one channel inside an ecosystem that includes wholesale and physical retail.

Retail. Mobile share over 60% on peak days. Inflation pressure is creating the "Lipstick Effect" pattern: consumers cut big-ticket durable goods, but continue to spend on small premium items. Beauty, wellness, and small accessory categories outperform.

Hospitality. Trust depends on visible operator content (chef-led TikTok and Instagram, named-staff stories) and verified review surfaces. AI-generated review skepticism is high. Google Business Profile completeness, real photos, and direct booking links matter more than review volume.

Services-as-product (coaching, courses, productized consulting). Founder visibility is required, not optional. The category is hit hardest by AI-generated content saturation. The differentiation is operator specificity: real client outcomes with named clients, real session footage, real published frameworks. Generic "transform your business" copy reads as AI-generated by default.

Implications for Operators in H2 2026.

Seven recommendations, grounded in the research above.