B2B buying journey: buyers pick a winner before they call you

By the time a buyer books the first sales call, the shortlist is ranked and the favourite usually wins. The selling now happens in places your sales team never sees.

People seated in a meeting room listening to a presentation
Photo: Jiří Sedláčej, CC BY-SA 4.0, via Wikimedia Commons

Most of the B2B buying journey now happens before a buyer talks to any seller. 6sense's 2025 survey of about 4,000 buyers found that first contact came, on average, 61% of the way through the purchase, that 94% of buying groups had already ranked their shortlist by then, and that the vendor ranked first went on to win 77% of the time. So the website, case studies, peer reputation and AI-search presence decide who gets the deal; the sales call mostly confirms it.

Key takeaways
  • 61% of the journey is done before a buyer first contacts a seller (6sense, 2025).
  • 94% of buying groups have ranked their shortlist by then, and the favourite wins 77% of deals.
  • 67% of B2B buyers prefer a rep-free experience, yet 69% want a rep to check AI-generated answers (Gartner).
  • Hidden buyers in finance and operations rarely meet sales but read thought leadership; finance sat in 46% of software decisions in 2026.
  • Website and sales must match: 69% of buyers found the two telling different stories.

This is not a new complaint, but the numbers have hardened and a new actor has arrived. In 2024 and 2025 the research pointed at peers and review sites. By 2026 it points at ChatGPT and AI agents too: Gartner, G2 and Wynter all found large shares of buyers starting their research in an AI assistant.

I spent years marketing B2B voice AI at Gnani.ai, where the deals were long and the buying groups large, and what follows is how I'd read the evidence and what I'd change because of it.

61%of the journey done at first seller contact (6sense, 2025)
77%of deals go to the buyer's pre-contact favourite (6sense, 2025)
67%of B2B buyers prefer a rep-free experience (Gartner, 2026)
84%of surveyed SaaS CMOs use AI to discover vendors (Wynter, 2026)

How much of the B2B buying journey is over before sales gets involved?

About three-fifths of it, on 6sense's measure. Its 2025 Buyer Experience Report, released on 12 November 2025, put the average point of first contact at 61% of the way through the buying process, down from 69% in earlier editions.

That shift looks like good news for sellers until you see what else the report found: buyers move earlier not because they want advice, but because economic pressure and AI questions push them to.

The story

The idea that buyers do most of the work alone has a birthday. Through 2011 and 2012, Google worked with CEB's Marketing Leadership Council to survey 1,500 business leaders who had recently been involved in major purchases for 22 large B2B organisations. In February 2013 Google's Corey Mull wrote up the result: on average, business buyers did not contact suppliers directly until 57% of the purchase process was complete.

The write-up opened with an IT buyer told by the CEO to fix freezing computers on a budget that could grow by only 10%. The buyer searched, read bloggers and forums, picked three or four vendors, invited them to present and took the lowest bid, all without a single salesperson's input. The "57%" slide went on to appear in B2B decks for a decade. Thirteen years later, 6sense's number is 61%.

Think with Google: B2B's Digital Evolution, Corey Mull (February 2013)

The shortlist is ranked before the first call

The numbers that matter more are about the shortlist. In the 2025 study, 94% of buying groups had ranked their shortlisted vendors by preference before contacting any of them. In practice that meant:

  • they contacted their top-ranked vendor first about 80% of the time;
  • they bought from that vendor 77% of the time;
  • buyers, not sellers, initiated the first contact in 79% of cases;
  • they evaluated about five vendors on average.
"Buyers are choosing a preliminary winner much earlier than they have in the past."Kerry Cunningham, head of research at 6sense, in the 2025 release
What B2B buyers had decided before first contact, 2025
See the data as a table
Buyers (%)
Ranked their shortlist before contacting any vendor94
Contacted their top-ranked vendor first80
Initiated first contact themselves79
Bought from their pre-contact favourite77
Share of the journey done at first contact61

Source: 6sense 2025 Buyer Experience Report, via Business Wire release and CustomerThink summary · Nearly 4,000 responses across North America, EMEA and APAC; 42% technology and 38% services companies.

Who answered the 6sense survey

The sample is worth knowing before you quote it. CustomerThink's breakdown of the report shows:

  • 46% of respondents in North America, 20% each in continental Europe and the UK and Ireland, and 14% in Asia-Pacific;
  • 42% in technology and 38% in services;
  • nearly half at companies with $10 million to $500 million in revenue.

India is inside "Asia-Pacific" at best. Treat the pattern as global and the decimals as American.

Buyer-led research: 61%With sellers: 39%First seller contactTriggerContractPeers and communitiesReview sitesVendor websitesAI assistants94% rank the shortlist hereFavourite wins 77% of dealsShares from 6sense 2025 Buyer Experience Report
Where the decision gets made. The sales team enters after the ranking is done, which is why the work before that line decides most deals.

Do B2B buyers even want to talk to a salesperson?

Most say they would rather not, until they need judgement they cannot get online. Gartner's survey of 632 B2B buyers, run in August and September 2024 and released on 25 June 2025, found that 61% prefer an overall rep-free buying experience. The next wave, 646 buyers surveyed in August and September 2025 and released in March 2026, put the figure at 67%, with 45% saying they had used AI in a recent purchase.

The same Gartner releases explain why. In the 2025 release, 73% of buyers said they actively avoid suppliers who send irrelevant outreach, and 69% reported inconsistencies between what a vendor's website said and what its sellers told them.

Gartner's Robert Blaisdell said mismatched information breeds mistrust. That second number is the one I'd put on a sales leader's wall: if your site and your reps tell different stories, the buyer believes neither.

The preference is not absolute. In May 2026 Gartner reported, from a survey of 645 buyers, that 69% turn to sales reps to validate AI-generated insights. Buyers do the research alone, then want a human to check whether the machine was right about fit, risk and price. The rep's job has moved from informing to verifying.

The Hall of Nations at Pragati Maidan, Delhi, a concrete space-frame exhibition building, photographed in 2008
The Hall of Nations at Pragati Maidan, Delhi, designed by Raj Rewal and built in 1971-72 for trade fairs. Trade fairs still create first impressions, but a buyer who walks up to a stall with intent has usually written the shortlist already.Photo: Ariel.huber, CC BY-SA 3.0, via Wikimedia Commons

Where do buyers do their research now?

Buyers now research in four places, and the AI share has grown faster than any channel in marketing memory:

  • peers and communities;
  • review sites such as G2;
  • vendor websites;
  • AI assistants such as ChatGPT and Perplexity.

Wynter surveyed 101 CMOs at mid-market B2B SaaS companies in January 2026 and found that 84% use AI tools or LLMs to discover vendors, up from 24% a year earlier and 0% in 2024. Sixty-eight per cent start their search in an AI tool before a search engine, and 65% start in peer communities.

SaaS CMOs who use AI tools to discover vendors
See the data as a table
Share of CMOs (%)
20240
202524
202684

Source: Wynter: How B2B SaaS CMOs buy software in 2026 · 2026 wave: 101 CMOs at B2B SaaS companies with $50M+ revenue, January 2026. Small, US-centric sample.

Review sites still edge out chatbots

G2's 2026 Buyer Behavior Report, published in July 2026 from a survey of more than 1,000 software buyers, puts the shift in context. Eighty-two per cent had sourced software recommendations from an AI chatbot in the previous 24 months. Yet when asked what most shaped the shortlist, review sites (38%) edged ahead of AI chatbots (37%).

The two are linked: AI assistants lean heavily on review sites, forums and comparison pages when they answer, which is why the domains AI cites matter so much. I list them in the most-cited domains in AI search.

Familiarity has risen with it. In Wynter's 2026 sample, 80% of CMOs arrived at the first sales call at least moderately familiar with the vendor, and 48% were very familiar, up from 22% in 2025. A first call is no longer an introduction. It is a check on whether the vendor matches the picture the buyer already built from other sources.

StudySample and dateHeadline finding
6sense Buyer Experience ReportAbout 4,000 buyers, global, released Nov 2025First contact at 61% of the journey; favourite wins 77%
Gartner sales survey632 buyers, Aug–Sep 202461% prefer rep-free buying; 73% avoid irrelevant outreach
Gartner sales survey646 buyers, Aug–Sep 202567% prefer rep-free buying; 45% used AI in a recent purchase
Wynter CMO survey101 SaaS CMOs, Jan 202684% use AI to discover vendors; 68% start there before search
G2 Buyer Behavior Report1,000+ software buyers, Jul 202682% got recommendations from AI chatbots; review sites 38% vs AI 37% as top shortlist source
Edelman-LinkedIn1,934 executives, Jun 2025Over 40% of deals stall on buying-group misalignment

Who else is deciding, and why can't sales reach them?

A buying decision is made by a group, and most of that group never meets your salesperson. 6sense's 2024 report put the average buying group at 11 people.

The 2025 Edelman-LinkedIn report, Invisible Influence, surveyed 1,934 executives and focused on hidden buyers: people in finance, operations and leadership who shape the decision without being the obvious target. It found that more than 40% of B2B deals stall because the buying group cannot align internally, and Mi3's summary of the report says about seven in ten hidden buyers have little or no contact with vendor sales teams.

Who your salesperson actually meetsTarget buyerschampion, user, ITHidden buyersFinanceOperationsLeadershipProcurementSales repdemo, callsblogs, reviews,AI answersAverage buying group: about 11 people (6sense, 2024)
The buying group as two rings. Sales reaches the inner ring; the outer ring forms its view from what it reads, which is why public content reaches people a demo never will.

What hidden buyers read

Those hidden buyers read. In the 2025 report, 63% of them spent more than an hour a week on thought leadership, almost the same as target buyers at 64%, and 55% used it when building vendor shortlists. Sixty-four per cent trusted it more than marketing materials or product sheets when judging a vendor's capability.

G2's 2026 data adds the money angle: finance was involved in 46% of software decisions, up from 31% a year earlier, and nearly half of buyers said a CFO had vetoed an approved deal. The CFO who kills your deal has probably never seen your demo, but may have read your blog.

Edelman-LinkedIn finding2024 report (Dec 2023 fieldwork, about 3,500 people)2025 report (1,934 executives)
Trust thought leadership over marketing materials to judge capability73%64% of hidden buyers
Thought leadership led them to a product they weren't considering75% of decision-makers and C-suiteNot reported
Spend an hour or more a week reading it52% of decision-makers, 54% of C-suite63% of hidden buyers, 64% of target buyers
Thought leadership made them question an existing supplier70% of C-suite, at least occasionallyNot reported
Use thought leadership when building shortlistsNot reported55% of hidden buyers
The Main Reading Room of the Library of Congress, with desks arranged in circles under a domed ceiling
The Library of Congress reading room. B2B buyers now do most of their reading before they meet a vendor, and hidden buyers in finance and operations do almost all of theirs that way.Photo: Carol M. Highsmith, Public domain, via Wikimedia Commons

How do AI assistants decide which vendors to recommend?

AI assistants recommend the vendors whose information is easiest to find, cite and trust, and early research shows the choice can be shifted by how pages are written. Two open papers make this concrete.

GEO: Generative Engine OptimizationPranjal Aggarwal et al. · KDD 2024

The authors built GEO-bench, a benchmark of queries across many domains with matched web sources, and tested nine ways of rewriting a source page to see which made a generative engine quote it more. Adding citations, quotations and statistics raised visibility by more than 40% across queries, and by up to 37% on Perplexity.ai in a live test. The gain was largest for weak pages: a source ranked fifth in search results gained 115% visibility with the "cite sources" rewrite, while keyword stuffing did little. The limit: engines change fast, and the visibility metric is the authors' own.

Read the paper (free)
What Is Your AI Agent Buying? Evaluation, Biases, Model Dependence, and Emerging Implications for Agentic E-CommerceAmine Allouah, Omar Besbes, Josué D. Figueroa, Yash Kanoria, Akshit Kumar · arXiv · 2025

The authors built ACES, a test marketplace, and let AI shopping agents choose among eight products per category across eight categories, with 1,000 market-share runs per category and randomised experiments on layout, labels, ratings and price. Agents concentrated demand on a few products, and the favourites changed sharply between model versions. A platform's "Overall Pick" endorsement lifted a product with a 10% baseline to 20–43% selection depending on the model, while a "sponsored" tag cut it. When sellers let an AI rewrite their product description for agents, market share rose by up to about 15 percentage points on average for some models. The limit: these are consumer products in a simulated store, not seven-figure B2B contracts.

Read the paper (free)

For B2B the lesson transfers in shape, not in size. Agents and assistants reward pages that state facts plainly, carry third-party endorsement and cite evidence, and they punish what looks like an ad.

G2's 2026 survey found 61% of buyers already use or plan to use AI agents in buying, with 51% using them to build shortlists, though only 9% would let an agent actually purchase within guardrails. If an agent is drafting the shortlist, your public pages are the pitch. I go into tactics in how to get cited by ChatGPT and Perplexity.

What does being the favourite before first contact actually win you?

Being the buyer's favourite at first contact is worth roughly three to four times the win rate of being one of the others. Use this to check what you just read: the calculator below uses 6sense's 2025 numbers: the favourite wins 77% of deals, and the remaining 23% is shared among the other shortlisted vendors, about four of them when buyers evaluate five. Plug in your own pipeline.

Wins from being the pre-contact favourite

At 100 opportunities and a 30% favourite share, the model gives about 27 wins. Raise the favourite share to 50% and it gives about 41. Few sales-training programmes would even promise a lift that size. The only lever that big is being known and preferred before the buyer reaches out, which is a marketing job.

What should a B2B company change because of this?

Move effort from chasing buyers to being findable and credible where they research, and make sure sales and the website say the same thing. Here is the order I'd work in.

  1. Fix the website first. Gartner's 69% inconsistency figure means the site is often wrong or vague. Put pricing logic, integrations, security answers and implementation time on public pages. Wynter's 2026 report makes the same point: publish verifiable proof, methods and sources.
  2. Write case studies a CFO can check. Named customer, baseline, result, time period. A hidden buyer in finance reads these, not your demo. My template is in how to write a B2B case study that sells.
  3. Earn third-party mentions. Reviews, analyst notes, community threads and comparison pages are what both humans and AI assistants trust. The ACES study shows an endorsement badge moves agents far more than a sponsored slot.
  4. Publish real thought leadership. Edelman's data says it reaches the hidden buyers sales never meets, but only if it says something specific. A founder writing weekly is the cheapest version; see a founder-led LinkedIn content system.
  5. Retrain sales to validate, not pitch. Buyers arrive with a favourite and an AI-made summary. The rep who can say where that summary is wrong, and prove it, wins the 23% and protects the 77%.
  6. Measure the before-contact stage. Track branded search, direct traffic, AI-referral visits and self-reported "how did you hear about us". Last-click attribution will credit the demo form for a decision made weeks earlier; B2B lead sources ranked by conversion shows why inbound converts better.
Rows of server racks inside a data centre in Amravati, India
Server racks at a data centre in Amravati. The AI assistants that now draft many shortlists answer from what they can find and cite on the open web.Photo: PiDatacenters, CC BY-SA 4.0, via Wikimedia Commons

Is the sales call dead, then?

No. The sales call has moved from the start of the decision to the end, and it now decides the close, not the shortlist. 6sense found 62% of buyers engaged sellers earlier because of economic uncertainty, and 58% did so to understand how vendors implement AI.

Gartner's May 2026 finding that 69% of buyers ask reps to validate AI-generated insights says the same: buyers want a human where the risk is. What has died is the cold call as a way to create demand from nothing, and the idea that a first meeting is where you make your case.

For Indian B2B companies selling to global buyers, this is an opening. Buyers will not fly to Bengaluru to evaluate you, but they will read your documentation, your case studies and what ChatGPT says about you, at no cost to either side. A small team with clear public proof can be the favourite before the first call. That is the kind of marketing I describe on my work page.

Questions people ask

What percentage of the B2B buying process is done before contacting sales?

6sense's 2025 report puts it at 61% on average, down from 69% in its earlier studies. Its 2024 release said buyers had largely set their purchase requirements before reaching out in 85% of cases.

Do B2B buyers prefer to buy without a salesperson?

Gartner found 61% preferred a rep-free experience in its survey fielded in 2024, and 67% in the 2025 wave. But 69% still want a rep to check AI-generated insights, so the preference is about the research phase, not the whole deal.

How many vendors do B2B buyers consider?

About five on average in 6sense's 2025 data, with four of five shortlist slots filled on day one. Wynter's smaller CMO survey found buyers often arrive at the first call already very familiar with the vendor they favour.

Does thought leadership really influence B2B purchases?

The Edelman-LinkedIn studies say yes: 73% of respondents in 2024 trusted it more than marketing materials, and 55% of hidden buyers in 2025 used it to build shortlists. Both studies are run by firms that benefit from the answer, so treat the size of the effect with care, but independent data on early shortlisting points the same way.

Is LinkedIn worth it for B2B marketing?

For reaching hidden buyers who never meet sales, the evidence says it can be, because that is where much B2B thought leadership is read. Whether paid ads there pay back depends on deal size and targeting; organic posts from a founder or expert are usually the cheaper first test.

Sources

  1. 6sense via Business Wire: The timeline for influencing B2B buyers is shrinking (2025)
  2. CustomerThink: 6sense study provides critical insights on B2B buyer behavior (2025)
  3. 6sense: 2024 Buyer Experience Report press release (2024)
  4. Gartner via MarketScreener: 61% of B2B buyers prefer a rep-free buying experience (2025)
  5. IT-Online: Gartner finds 67% of B2B buyers prefer a rep-free experience (2026)
  6. Gartner: 69% of B2B buyers turn to sales reps to validate AI-generated insights (2026)
  7. Wynter: How B2B SaaS CMOs buy software in 2026 (2026)
  8. G2: AI is reshaping how B2B software deals are won and lost (2026)
  9. O'Dwyer's: 2024 Edelman-LinkedIn B2B Thought Leadership Impact Report (2024)
  10. Marketing-Interactive: Over 40% of B2B deals stall due to hidden buyers (2025)
  11. AMI / Mi3: Hidden buyers stalling 40% of B2B deals (2025)
  12. Think with Google: B2B's Digital Evolution, CEB Marketing Leadership Council research written up by Corey Mull (2013)
  13. Aggarwal et al.: GEO, Generative Engine Optimization, KDD (2024)
  14. Allouah et al.: What Is Your AI Agent Buying? (2025)
B2B buying journey6senseGartnerEdelmanLinkedInbuying groupAI searchthought leadership
Sapta

Sapta (Saptarshi Sarkar) is a mechanical engineer turned marketer in Bengaluru and a solo touring rider. He writes the seven Saptaverse blogs. Work with him or read more of Marketblog.

Keep reading in Marketblog

From the other blogs