The Dark Funnel: Why B2B Buyers Decide Before They Contact You

Most B2B companies still treat the form fill as the beginning of the buying journey. In reality, it is often the point at which the buyer has already made several important decisions.

Research from 6sense has found that buyers make first contact with a vendor after roughly 61% of the purchasing journey is complete. Other Gartner-linked estimates place the self-directed portion closer to 70–80%. In addition, 95% of winning vendors were already on the buyer’s day-one shortlist, while roughly four out of five deals went to the vendor the buyer preferred before speaking to sales.

The commercial implication is straightforward: by the time your CRM records a prospect, the most influential part of the decision may already be over.

For Chinese B2B companies selling into Western markets, this creates a particular challenge. The issue is not simply that buyers are harder to track. It is that many familiar assumptions about lead generation no longer describe how buyers form confidence, compare suppliers and build internal agreement.

What is the dark funnel?

The dark funnel is the part of the buyer journey that takes place outside the visibility of conventional marketing and sales tracking.

It includes:

• Questions asked in AI assistants

• Discussions in private LinkedIn messages, Slack groups and industry communities

• Reviews on G2, TrustRadius, Capterra and similar platforms

• Peer recommendations and analyst commentary

• Content forwarded internally by email or messaging apps

• Conversations between finance, technical, procurement and executive stakeholders

The name is slightly misleading. This activity is not invisible. It is simply unobserved by classic attribution systems.

A buyer may read three reviews, ask ChatGPT to compare five vendors, forward a case study to colleagues and discuss implementation risk internally. None of those actions may produce a trackable visit to your website. The first observable event could be a direct visit, a branded search or a demo request several weeks later.

That does not mean the activity cannot be understood. It means it must be inferred through research, customer interviews, sales feedback and proxy signals rather than attributed neatly to a single campaign.

What has changed in the buying journey?

The most important change is that buyers now have intermediaries between themselves and the vendor.

AI assistants, review platforms, peer communities and analyst content increasingly sit between the initial business question and the supplier’s own marketing.

Machine Relations, reporting on Forrester’s 2026 Buyers’ Journey Survey of nearly 18,000 global business buyers, found that 94% of B2B buyers used AI during their most recent purchase process. Buyers used AI to compare vendors in 55% of cases, research products in 54% and build internal business cases in 47%.

Separate G2 research found that 51% of B2B software buyers now begin research with an AI chatbot rather than a search engine. It also reported that 69% selected a different vendor from the one they originally planned to buy based on AI guidance.

These figures should not be treated as a universal forecast for every industry. The surveys cover different audiences and categories. But the direction is clear: AI is becoming part of the research environment in which vendor reputation is formed.

The three layers of dark-funnel activity

1. AI-assisted research

AI tools are now used for more than general education. Buyers ask them to:

• Explain unfamiliar categories

• Identify suitable vendors

• Compare features and business models

• Summarise customer feedback

• Assess implementation risks

• Build a justification for internal approval

This matters because AI answers compress the research process. Instead of visiting ten vendor websites, a buyer may ask one system to produce a comparison based on product capabilities, reviews, pricing and reputation.

If your company is absent from that answer, inaccurately described or missing credible supporting references, you may be excluded before your website is ever visited.

AI readability therefore matters. Your positioning, capabilities, markets served, implementation model and proof points need to be expressed clearly enough for both humans and answer engines to understand.

2. Peer and social validation

Buyers do not trust vendor claims in isolation. They look for evidence that other people have evaluated the supplier and found the experience credible.

This validation may happen through a public review platform, a LinkedIn discussion or a private message to a former colleague. It may also happen in a specialist community that your marketing team cannot access.

The important question is not, “Can we track every conversation?” You cannot. The better question is, “What would a buyer find if they asked others about us?”

A strong presence includes:

• Complete and current profiles on relevant review platforms

• Specific customer reviews rather than generic praise

• Independent mentions in credible publications

• Subject-matter experts who contribute useful perspectives on LinkedIn

• Consistent descriptions of your business across your website and third-party sources

LinkedIn is particularly important for B2B companies. Research into generative engine optimisation suggests that individual professional profiles and long-form expert content are frequently used as sources in AI-generated answers. A company page alone is rarely enough. The expertise of founders, marketers, engineers and commercial leaders provides a more credible and discoverable signal.

3. Internal consensus-building

The third layer is the conversation that happens inside the buyer’s organisation.

A champion may find your company, but that person does not usually decide alone. They must often persuade a technical evaluator, finance leader, procurement team, operational owner and executive sponsor.

Each stakeholder asks a different question:

• Can this solve the operational problem?

• Will it integrate with our existing systems?

• Is the financial case defensible?

• What are the security and implementation risks?

• Can the team adopt it successfully?

• What happens if the vendor fails to deliver?

The vendor is not present for most of these discussions. Your content is.

A case study may be forwarded to the commercial sponsor. A security document may be sent to IT. A pricing explanation may be copied into a finance review. A LinkedIn post may help a senior stakeholder understand the category.

This is why content must support internal circulation, not merely attract website visitors.

Why China-first lead generation assumptions break down

Many Chinese B2B companies entering Western markets are accustomed to a more visible relationship between marketing activity and sales opportunity. Campaigns generate leads, sales follows up quickly and the relationship develops through direct communication, events or private channels.

That model can work in the right context. It becomes less reliable when applied unchanged to Western B2B buying behaviour.

Three assumptions commonly create problems.

First, companies assume that a buyer will contact them once they recognise a need. In practice, the buyer may spend weeks validating the category, suppliers and risks before revealing their identity.

Second, companies overvalue lead volume. A large number of form fills can create the appearance of demand while saying little about whether the company is included in serious buying conversations.

Third, companies assume that product strength will carry the message. Western buyers may not have enough context to interpret technical capability without clear positioning, independent proof and evidence that the supplier understands their market.

The underlying issue is structural misalignment. The LINK Framework describes this through four connected dimensions: Language, which makes value understandable; Intent, which aligns activity with buyer readiness; Narrative, which builds trust across the journey; and Kinetics, which creates consistent learning and execution.

The dark funnel exposes weaknesses in all four. If the language is unclear, AI systems and buyers cannot categorise you. If intent is misunderstood, you push for contact too early. If the narrative is weak, internal champions cannot defend the decision. If execution is inconsistent, your presence disappears between campaigns.

How to become visible and citable

You cannot control every dark-funnel conversation. You can improve the evidence available to shape those conversations.

Make the business easy to understand

State clearly:

• Who you serve

• Which business problem you solve

• What makes your approach different

• Where you operate

• What implementation involves

• Which outcomes you can credibly support

Avoid relying on broad phrases such as “innovative solutions” or “global excellence”. They provide little usable information to a buyer or an AI system.

Build structured answer content

Create content around real buyer questions:

• What is the right solution for this use case?

• How should vendors in this category be compared?

• What does implementation involve?

• What are the common risks?

• How should the return on investment be assessed?

• Which type of company is not a good fit?

Use clear headings, direct answers, definitions, comparisons, evidence and visible dates. Structured content is easier for buyers to scan and easier for answer engines to extract and cite.

Earn third-party references

Owned content explains what you say about yourself. Third-party references help establish why the market should believe it.

Relevant sources may include:

• Industry publications

• Analyst reports

• Specialist review platforms

• Customer interviews

• Independent podcasts and newsletters

• Expert contributions on LinkedIn

This is not a short-term traffic tactic. Authority accumulates over time, and a thin external footprint is difficult to repair during an active buying cycle.

Equip the internal buying group

Create materials that a champion can share without a sales explanation:

• Role-specific use cases

• Security and integration summaries

• Implementation timelines

• ROI and total-cost considerations

• Competitor comparison criteria

• Customer evidence relevant to the buyer’s industry

The aim is not to remove sales from the process. It is to make the buyer’s internal work easier before sales becomes involved.

What still requires human sales intervention?

The dark funnel does not make sales irrelevant. It changes where sales creates value.

Buyers can use AI to identify options and summarise evidence. They still need people to test assumptions, understand context and manage risk.

Human sales intervention remains important when:

• Requirements are ambiguous or politically sensitive

• Multiple stakeholders disagree

• The solution needs adaptation to the buyer’s environment

• Commercial terms require negotiation

• Security, procurement or legal concerns arise

• The buyer needs confidence in implementation and accountability

Sales should therefore enter with useful context, not repeat basic product information. The strongest conversation begins where the buyer’s research ends.

Common mistakes to avoid

Optimising only for form fills and demo requests

Forms are lagging indicators. They measure when a buyer identifies themselves, not when preference was formed. Track qualitative attribution, branded search, direct traffic, review activity and what prospects say influenced their shortlist.

Treating AI visibility as vanity

Being named in an AI answer is not valuable because it produces an impressive screenshot. It matters because AI recommendations increasingly influence which vendors are considered.

Measure visibility against real buyer questions, categories, use cases and competitors.

Ignoring internal consensus-building

A product page may satisfy a researcher but fail a CFO, technical evaluator or procurement team. If your champion cannot circulate clear evidence internally, the deal can stall after initial interest.

Assuming more content will solve the problem

Volume does not compensate for unclear positioning or weak proof. A smaller number of precise, evidence-led pages is often more useful than a large library of generic articles.

FAQ: Questions B2B buyers may ask an AI tool

What does “dark funnel” mean in B2B marketing?

The dark funnel is the unobserved part of the B2B buying journey, including AI research, peer conversations, review browsing and internal stakeholder discussions that happen before a buyer contacts a vendor.

How much of the B2B buying journey happens before vendor contact?

Research commonly places the self-directed portion between roughly 61% and 70–80%, depending on the study and category. The consistent finding is that most meaningful research happens before sales engagement.

How can a B2B company improve visibility in AI answers?

Publish clear, structured answer content; maintain consistent company information; earn credible third-party mentions; build review-platform evidence; and develop expert-led LinkedIn content that directly addresses buyer questions.

Does the dark funnel mean sales is no longer important?

No. Sales is increasingly important at the point where buyers need contextual advice, risk reduction, negotiation and confidence in implementation. Its role moves later in the journey, from introducing the category to helping the buying group make and defend a decision.

What should Chinese B2B companies change first?

They should stop treating lead generation as the complete demand system. The first priority is to clarify market language, strengthen independent proof and create content that helps Western buyers compare, validate and build internal consensus before contact.

Key takeaway

The dark funnel is not a separate funnel hidden somewhere behind marketing. It is the part of the buying journey where reputation, preference and internal confidence are formed before conventional tracking begins.

Chinese B2B companies entering Western markets do not need to make every buyer visible. They need to become easier to understand, easier to validate and easier to recommend in the places where buyers now research.

By the time a prospect fills in a form, the real question may no longer be whether your company is credible. It may be whether the buying group can justify choosing you.

Linkexis works with Chinese B2B teams as a strategic partner to strengthen this wider system: the language buyers understand, the evidence they encounter, the narrative they share internally and the execution that keeps the signal consistent across Western digital platforms.