Buying has moved further from the sales rep. Gartner’s March 2026 survey of 646 B2B buyers found 67% prefer a rep-free experience and 45% used AI during a recent purchase. G2’s March 2026 study (n=1,076) found 51% of B2B software buyers now start research in an AI chatbot, 71% rely on chatbots somewhere in the process, and a third bought from a vendor a chatbot surfaced that they had never heard of. Meanwhile the room got bigger: buying groups now average nine members, two-thirds of committees have six or more stakeholders, and the average buying cycle has compressed to about seven months. If your ICP document doesn’t say who is in that room, what they are researching, and what you can prove, your demand system is guessing.
How to use
One page per primary segment. Copy the sections below into your wiki, Notion, or CRM record, fill them with your real data, and assign an owner. Review every 90 days — the committee map and channel reality both shift.
Segment name
One sentence: segment label plus the trigger that makes them buy now (e.g., “Series B–C B2B SaaS CFOs who just missed a renewal target”).
Why this field matters: Shortlists now form before you are in the room — 51% of software buyers start with an AI chatbot, and one-third purchased from a vendor they had never heard of beforehand. A named segment with a trigger gives your demand system something concrete to target and measure.
Firmographics / context — fit vs. intent
Two columns, deliberately separate:
- Fit (firmographics): revenue band, employee count, industry, tech stack, geography, current vendor landscape.
- Intent (behavioral): active research signals — category, competitor, comparison, and pricing pageviews on software review platforms; downloads of problem-adjacent content; form fills on comparison queries.
Why this field matters: Fit says “could buy”; intent says “is researching now.” In the INFUSE + G2 study across 11 programs, accounts showing both intent and campaign engagement produced 20.9% more leads per account, 93.3% higher multi-touch engagement, and were 1.6x more likely to sit in the decision stage than accounts reached by activation alone. Keep the columns separate or SDRs will conflate a good-looking company with an in-market one.
Problems they already feel (signal language)
Capture the exact phrases buyers use — from sales calls, support tickets, community posts, and chatbot queries — plus the trigger events behind them.
Why this field matters: Buyers self-serve research, and they search in problem language, not product language. Gartner finds 99% of B2B purchases are driven by organizational changes, so anchor each problem to a trigger (new mandate, missed target, audit finding, leadership change). Problem phrases become your outbound openers, your landing page headlines, and the content chatbots can quote.
Disqualifiers / negative ICP
A hard list: company attributes and situations that mean “do not spend time or budget here” (e.g., below minimum revenue, reseller-dominated motion, procurement requires sole-source multi-year contracts, already in an active renewal dispute with your category).
Why this field matters: With average cycles compressed to ~7 months, time spent on bad fits is time not spent on a nine-member buying group. Negative ICP also keeps your scoring honest: a company can fail fit but show intent noise, and the system needs a rule for that, not a judgment call per lead.
Buying committee roles
Map at least these four per segment, plus any role-specific blocker:
- Champion: the internal seller who pushes the project forward; usually the person who feels the problem most acutely.
- Economic buyer: owns budget approval; cares about payback and risk, not features.
- Technical evaluator: defines requirements, runs the evaluation, often the security or architecture gate.
- Blocker: the skeptic — incumbent-vendor loyalist, integration owner, or the stakeholder a previous deal burned.
Why this field matters: Consensus, not champions, closes deals — two-thirds of B2B committees now have six or more stakeholders, and a single engaged contact is no longer account momentum. In the INFUSE + G2 data, 85–99% of intent leads sat at Manager level or above, and VP-level engagement ran up to 46% higher in intent-aligned accounts. List who you can reach, who you can’t, and who can kill the deal.
Proof you can show today
An evidence inventory — only assets that already exist: named case studies (with permission), industry benchmarks, audit or assessment results, an ROI calculator, customer quotes with role and company type, third-party review profiles.
Why this field matters: Chatbot-mediated buying rewards the verifiable. G2 found 83% of buyers feel more confident in a chatbot-assisted decision and 85% think more highly of a vendor a chatbot mentions — but a chatbot can only cite what it can check. If your ICP page lists proof that doesn’t exist yet, the first job is producing it, not prospecting.
Offers that fit
Match offers to problem stage, not product stage: low-commitment assets that work without a rep (benchmark report, template, calculator, audit tool) and rep-assisted steps for the late committee (demo with the technical evaluator, security review, TCO walkthrough).
Why this field matters: Two-thirds of buyers prefer a rep-free experience, and buyers consume roughly nine content pieces across the journey. Offers are the only thing your content can ask for — make the first ask cheap, the second ask role-specific.
Channels that earn attention
Per segment, list 2–4 channels with the job each does. 2026 reality checks:
- Email: the 2025 global inbox placement rate was 87.2%, but B2B placement fell 1.3% at hosting platforms and 5.8% at corporate filters; Microsoft is the hardest (77.4% vs Gmail’s 89.8%). Since February 2024, senders of 5,000+ messages/day to Gmail must authenticate (SPF/DKIM/DMARC), offer one-click unsubscribe, and stay under a spam threshold — and Validity reports triggered sends earn 10–15x more revenue per email than broadcasts.
- AI chatbots and answer engines: 51% of buyers start here; ChatGPT dominates at 63% — citable, dated content on your own domain is the requirement, and G2 is the second-leading shortlist influence after GenAI tools.
- Software review platforms: comparison pageviews are one of the strongest responsiveness predictors in the INFUSE + G2 data (competitor pageviews alone correlate with up to 41% more leads).
- LinkedIn: useful for reaching named roles in the committee, not for volume spray.
Sample segment (generic example — not real company data)
- Segment name: Mid-market B2B SaaS COOs (150–1,000 employees) whose net revenue retention dropped two quarters straight.
- Firmographics: $10–50M ARR; product-led sales motion; sales-led onboarding; no renewal automation. Intent: 2+ category or competitor pages viewed on G2 in the last 30 days; downloaded a churn-benchmark asset.
- Problems they feel: “Renewals slipped again this quarter”; “CS is drowning in QBR prep”; “usage goes flat 60 days after onboarding”; “finance keeps asking why expansion isn’t happening.”
- Disqualifiers: under $5M ARR; pure self-serve with no CS team; active renewal dispute; procurement requires multi-year sole-source contracts.
- Committee: COO (champion), CFO (economic buyer), Head of Customer Success (technical evaluator), CIO (blocker — integration risk).
- Proof: two named case studies in adjacent industries, a 12-month retention benchmark, an ROI calculator.
- Offers: retention benchmark report (self-serve), QBR-in-a-box template, 30-minute account audit (rep-assisted).
- Channels: email with benchmark from a verified sending domain; G2 profile with comparison content; LinkedIn posts pointing to the benchmark.
Done when
- Sales and marketing reject a bad-fit lead using the same disqualifier language — test it: run five recent leads past both teams and confirm agreement.
- Every role in the committee column has an owner and at least one asset matched to it.
- An SDR can classify a new lead as fit, intent, or neither in under 30 seconds.
- The page has an owner and a review date within 90 days, and it has been updated once since your last AI-research or committee change.
Citations
