The B2B buying process is shifting from human-led research to AI-mediated interpretation, and the shift is accelerating.
The average B2B buying cycle compressed from 11.3 months in 2024 to 10.1 months in 2025 (6sense, 2025) – not because buyers are doing less research, but because AI tools have made them faster. The self-directed research phase moved from a 70/30 split in 2024 to a 60/40 split in 2025: buyers engage slightly earlier, but the shortlist is already formed when they do.
Most influence in B2B buying now happens earlier during the phase vendors rarely see.
Direct answer
AI changes the B2B buying process by taking over early research, comparison, and option reduction before buyers engage with vendors.
Instead of manually gathering information, buyers rely on AI systems to interpret available signals, reduce uncertainty, and form shortlists. As a result, most influence is established before sales conversations begin and companies that lack consistent, attributable authority signals are excluded before they know they were being evaluated.
Position statement
The buying process has not become shorter. It has become front-loaded. Influence is now concentrated earlier, before vendors are aware a decision is forming.
Key takeaways
- 61% of the buying journey occurs before first sales contact. – 6Sense2025
- The average buying cycle compressed from 11.3 to 10.1 months in a single year (6sense, 2025)
- 81% of buyers have a preferred vendor at the time of first contact (6sense, 2024)
- 80% of B2B deals are won by the vendor preferred before first contact (6sense, 2025)
- 80% of buyers trust AI tools at least sometimes; 20% report reduced confidence due to AI inaccuracies (TrustRadius, 2025 / Forrester, 2025)
GenAI chatbots are now the #1 source influencing vendor shortlists – ahead of vendor websites and peer recommendations (G2, 2025)
What the first 70% of the B2B buying process used to look like
Traditionally, buyers spent the early stages searching for information, reading vendor websites, consuming articles and reports, and consulting peers. This work was manual, fragmented, and time-consuming.
Because research was slower, buyers contacted vendors sooner. Sales conversations helped shape understanding, not just confirm decisions. Vendors had meaningful access to the early-stage process. That access is narrowing. Buyers and AI systems increasingly form trust and preference before sales engagement.
The B2B buying journey has undergone a structural shift. Approximately 60% of the process is now self-directed – buyers identify approaches, solution categories, and vendors through AI-assisted research, peer input, and expert content before any direct vendor contact occurs. This is not a behavioral preference. It is the new default architecture of how decisions form.
The process typically begins with one person or a small team. They frame the problem, estimate the cost of inaction, and start building a preliminary longlist – using AI systems, analyst content, LinkedIn, and external references. An initial shortlist is often formed before the full buying group is involved. It is later reviewed by IT, Finance, Procurement, Security, and business owners, each of whom may add or remove vendors based on what they independently find. By the time a buying group convenes, the market has already been interpreted. Vendors not present in that interpretation are not debated. They are simply absent.
Preference follows the same pattern. During the Preference Building phase, buyers revisit expert profiles, LinkedIn activity, case studies, and analyst materials. Not to discover new options, but to rank those already on the shortlist and ask: “Do we want to work with these people? Can we defend this choice internally?” Preference becomes visible to the vendor only at the moment of first contact, through a demo request, an assessment, a pricing inquiry, or a request to speak with a named expert. By then, the ranking already exists.
This is Battle 1: Win Early Preference – the primary marketing objective in the self-directed phase. Marketing’s responsibility is no longer to support sales after engagement begins. It is to ensure the company is discovered, shortlisted, and preferred before the first sales conversation takes place. Sales inherits a market interpretation it did not create. Its role is to validate, contextualise, and strengthen a preference that is already forming, not to build it from scratch.
How AI reshapes the early stages of the B2B buying process
AI has become the starting point. Buyers increasingly begin by asking AI systems – ChatGPT, Gemini, Perplexity, Copilot, how to evaluate a problem or category. AI answers by collecting available information, summarising perspectives, highlighting common options, and reducing complexity. This happens before buyers define a shortlist themselves.
The result is compression, not elimination. The average B2B buying cycle dropped from 11.3 months to 10.1 months in a single year, a reduction of more than six weeks driven by faster AI-enabled research (6sense, 2025). What previously took weeks of manual comparison can now happen in a single session.
Faster evaluation shifts influence upstream – into the phase vendors cannot directly observe.
How AI changes decision formation
AI does not explore broadly. It narrows.
Vendors are removed when their role is unclear, their expertise is not attributable, their signals are inconsistent across sources, or their presence relies primarily on promotion. Exclusion happens silently and categorically – before a buyer forms a conscious shortlist.
By the time buyers speak with vendors, they are often validating a pre-existing preference. 81% of buyers already have a preferred vendor at the time of first contact (6sense, 2024). 80% of deals are won by the vendor preferred before first contact (6sense, 2025). Sales conversations confirm direction. They rarely reverse it. Exclusion happens silently.
What the new B2B buying process means for sales and marketing
Sales is not becoming irrelevant. It is becoming downstream. If sales enters after shortlist formation, its effectiveness depends entirely on whether the company was included in that shortlist. Relationships influence final selection – but they cannot compensate for early-stage exclusion.
Marketing now shapes interpretation, trust, and perceived relevance before any contact occurs. This is a structural shift: marketing no longer supports sales only at the point of engagement. It determines whether sales has anything to work with.
80% of buyers trust AI tools at least sometimes (TrustRadius, 2025). But 20% report reduced confidence due to AI inaccuracies (Forrester, 2025). This tension defines the strategic opportunity: companies whose Authority Signals are accurate, consistent, and attributable become the sources AI cites with confidence and the vendors buyers trust before speaking to anyone.
What companies misunderstand about the first 70% of the B2B buying process
Four common assumptions are inaccurate:
- That AI replaces sales. It does not. AI reallocates influence upstream; human judgment still closes deals.
- That AI generates leads. It does not. AI filters options; companies either survive that filter or do not.
- That the first 70% can be influenced through campaigns. It cannot. Campaign signals are promotional and non-attributable – the categories AI systems systematically deprioritise.
- The first 70% is influenced through Authority Signals: consistent, attributable, publicly accessible expertise that AI can interpret, cite, and recommend.
How can marketing influence an AI-generated shortlist before the company sees a lead?
AI Visibility – the ability of a B2B company to be clearly identified, compared, and referenced by AI systems during early-stage buying research is the first requirement. It is not sufficient on its own.
Companies must also ensure their expertise, positioning, and perspectives are consistently interpreted across multiple channels and sources. A single well-optimised page does not create the recurring signal pattern AI requires to justify inclusion.
HiFuture refers to the full solution as Authority Orchestration™: the strategic discipline of designing, connecting, and activating authority signals as a consistent ecosystem across the buying journey, so AI systems can correctly interpret, cite, and recommend the company and buyers can trust it.
Within this model:
- Company Authority Signals establish structural clarity – what the company does, who it serves, in which decision context it belongs
- Human Authority Signals establish attribution – named experts whose consistent perspectives AI can trace across platforms
- Third-Party Authority Signals establish external reinforcement – analyst mentions, earned PR, and citations that confirm the same interpretation from independent sources
When these signals are fragmented – each team producing signals in isolation – the result is Authority Fragmentation: the structural cause of silent exclusion during the first 70% of the B2B buying process.
Executive implication
The strategic question is no longer: “How do we influence buyers during sales conversations?”
It is: “Are we shaping interpretation during the first 70% of the buying process – before sales is involved?”
Buyers complete 70–80% of their evaluation before contacting sales (Forrester, 2025). The companies that consistently appear on the Day-One Shortlist are not those that engage earlier in the sales cycle. They are those whose Authority Signals are present, consistent, and AI-readable before a buyer’s first AI-assisted query.
If the answer is no, influence is already lost.
Sources
- Forrester, B2B Buying Study, 2025 – https://www.forrester.com/predictions/b2b-2025/
- 6sense, B2B Buyer Experience Report, 2025 – https://6sense.com/science-of-b2b/buyer-experience-report-2025/
- 6sense, B2B Buyer Experience Report, 2024 – https://6sense.com/science-of-b2b/2024-buyer-experience-report/
- TrustRadius, Buyer Research, 2025 – https://solutions.trustradius.com/vendor-blog/bridging-the-trust-gap-b2b-tech-buying-in-the-age-of-ai/
- G2, Buyer Behavior Report, 2025 – https://learn.g2.com/2025-g2-buyer-behavior-report
