# The decision was made before you arrived.

6sense asked 4,611 B2B buyers how they choose. The favourite going in won 82% of the time, and AI use inside the evaluation went down.

Published by Fiction Tribe, 2026-11-06. Canonical: https://velocity.fictiontribe.com/field-notes/decision-made-before-you-arrived

Key figure: 82% of deals go to the vendor favoured before first contact.

## The short version

- In a survey of 4,611 B2B buyers, the vendor they favoured before contacting any seller won 82% of the time.
- 92% of winners were on the first shortlist. A vendor missing from it had under a 10% chance.
- AI use inside evaluations fell from 94% to 81% in a year. Only 21% of buyers say an LLM shaped their view before first contact.
- The budget that matters is the one that runs between buying cycles, and almost nobody has a line for it.

We have sat in a lot of pitches that felt close. Most of them were not. The buyer walked in with a favourite, and the next ten weeks were a polite way of confirming it.

Now there is a number for that feeling. 6sense surveyed 4,611 recent B2B buyers across North America, Europe and Asia Pacific and published the results on 2 November. The vendor a buyer preferred before speaking to any seller went on to win 82% of the time. Ninety-two percent of eventual winners were on the very first shortlist. If you were not on that list, your odds were under one in ten.

Read that as a marketer and it stings. The demo, the proposal, the case study deck, the nurture sequence: for most deals, all of it happens after the outcome is already likely.

## When do B2B buyers actually decide?

Before you know they exist. The average buying cycle in the study ran 10.3 months, and about 60% of that time was spent in selection. But selection is mostly a buyer checking a preference they already hold. They have been through six to eight evaluations in the category before. They know the names.

> **82%** of deals go to the vendor the buyer favoured before first contact. 6sense, 4,611 buyers, 2026

What they do between purchases is the interesting part. Ninety-six percent say they keep up with vendors between cycles. The most common way is the plainest one: they look at vendor websites, cited by 36%. And 71% say their opinion of a vendor changed between one cycle and the next.

So preference is formed in the quiet months. It moves. And it moves in places no attribution tool is watching.

## Is AI search deciding who gets shortlisted?

Less than the last eighteen months of conference talks would suggest. This is the figure we did not expect. The share of buyers using an LLM somewhere in their evaluation fell, from 94% last year to 81% this year. Only 26% use one to discover vendors at all. Asked what shaped their view of vendors before first contact, 21% named an LLM. Thirty-four percent named what they already knew. Thirty-two percent named conversations inside their own company.

> The whole industry is optimising for the influence that ranks third.

We sell AI visibility audits, so we have every reason to wish that number were higher. It is not. What a model says about you still matters, because it is the one input you can read word for word and correct. But it is an input to a preference, and the preference is built mostly from memory and from colleagues.

## How far should you trust this study?

Hold one limit in plain view. 6sense sells intent data. "The decision happens before you can see it" is the exact problem its product claims to solve, so the framing is not neutral, and the fieldwork dates are not published.

The falling LLM number is why we take it seriously anyway. A vendor in this category has no reason to invent a figure that makes AI look less important. When a source reports something against its own interest, that is usually the part to believe.

## What should a marketing team do about it?

Three things, in order of how much they cost.

**Find out what buyers already think.** Ask the last ten people who bought from you, and the last ten who did not, when they first formed a view. Then ask the AI engines the same question your buyers ask and write down the answer verbatim. You are looking for the preference that exists before the cycle starts.

**Treat your website as the between-cycle channel.** It is where 36% of buyers go to keep up. Most B2B sites are built to convert someone in-market this quarter. Very few give a returning visitor a reason to notice what changed since they last looked.

> **10 months** is the average buying cycle. The preference that wins it was formed before the clock started.

**Give the quiet months a budget line.** Right now the money follows the ten weeks you can measure. The decision follows the ten months you cannot. A point of view published on a schedule, work shown as it ships, a founder who says something specific in public: none of it attributes cleanly, and all of it is what a buyer remembers when they write the first shortlist.

In-cycle marketing still has a job. It closes the deals you were already going to win and stops you losing them on a technicality. But it is a closing cost. The cause sits earlier, in a room you were never invited into, and the only way to be in that room is to have been worth remembering.

## Questions people ask

### When do B2B buyers decide which vendor to choose?

Mostly before they contact any seller. In 6sense's 2026 survey of 4,611 buyers, the vendor favoured before first contact won 82% of the time and 92% of winners were on the initial shortlist.

### How much do AI tools like ChatGPT influence B2B vendor selection?

Less than existing knowledge or internal discussion. Only 21% of buyers cited an LLM as an influence on their view of vendors before first contact, against 34% for existing knowledge and 32% for internal discussion. LLM use in evaluations fell from 94% to 81% in a year.

### What should marketers do if deals are decided before the buying cycle?

Fund presence between buying cycles. 96% of buyers keep up with vendors between purchases, most often through vendor websites, so a steady published point of view and a site that rewards return visits do more than in-cycle campaigns.

## Sources

1. [Vendors need to be known and preferred before the buying journey begins](https://6sense.com/news/6sense-research-finds-that-vendors-need-to-be-known-and-preferred-before-the-buying-journey-begins), 6sense, 2026-11-02. Survey of 4,611 B2B buyers.
2. [6sense: vendors need to be known and preferred before the buying journey begins](https://www.demandgenreport.com/industry-news/news-brief/6sense-vendors-need-to-be-known-and-preferred-before-the-buying-journey-begins/55214/), Demand Gen Report, 2026-11-03. Coverage of the same study.
