How B2B buyers build a vendor shortlist before they contact a seller

This guide is for B2B sellers, marketers and founders who want to be on more shortlists, and for buyers who have to build one. It explains how shortlists form, where buyers look, who shapes the list, why vendors get cut before anyone calls, and a test you can run this week to see where you stand. For the wider statistics on buying groups and timing, see B2B buying process statistics.

How a shortlist forms

Shortlists don't follow a fixed sequence, but most go through the same six steps, and sellers are absent from the first four.

1. A trigger and a Day 1 list. Something forces the decision: a contract ending, a failure, growth, a new leader. Bain calls the vendors a buyer has in mind at the start the "Day 1 list", and its June 2026 article on likelihood to buy says that a vendor not on it is "not really in the mix".
2. Must-haves. The group agrees what the purchase has to do and what rules a vendor out: price range, integrations, security standards, contract terms, size of vendor.
3. Research without sellers. Each member researches on their own: AI chatbots, review sites, analysts, peers, forums and vendors' websites.
4. Elimination. Vendors that fail a must-have go. So do vendors whose answers can't be found or checked without a sales call.
5. Ranking. The remaining vendors are put in order. In 6sense's 2025 Buyer Experience study of more than 4,000 buyers, 94% of buying groups had ranked their preferred vendors before first contact with any seller.
6. First contact and validation. Sellers are brought in to confirm fit, price and implementation. 6sense found the split between research and seller engagement moved from 70/30 to 60/40 in 2025, and buyers purchased from their preliminary favorite 77% of the time.

The practical meaning: a seller's first conversation is usually with a buyer who has a favorite and a list of questions, and the questions decide whether the favorite holds.

Where buyers look

G2 asked 1,076 B2B software decision makers in North America, EMEA and APAC (March 2026) which sources influence their shortlists. G2 runs a software review site, so read its findings with that interest in mind.

In the same survey, 51% of buyers said they start research with a chatbot more often than with Google, and 69% said an AI chatbot had led them to choose a different vendor than they had planned. 6sense found 94% of its buyers used large language models in buying, for example to summarize reviews. Whatever the exact share in your market, an assistant now answers some of the questions your website used to.

Who shapes the list

Bain distinguishes target buyers (users and product owners who know the problem) from hidden buyers, "process owners or fiduciaries" in procurement, finance and IT, and says recent LinkedIn and Bain analysis found hidden buyers have half of the influence over that early decision. Its survey of 750 buyers in two enterprise software categories (February 2026) measured five decision factors: meeting current and future needs, value relative to total cost, being a safe choice that can be defended internally, confidence in implementation, and ease of working with the vendor.

Why vendors get cut before anyone calls

This is guidance from practice, not survey data. The usual reasons a vendor drops off a list without ever hearing about it:

• No price. If two vendors publish prices and one says "contact us", the one that needs a call is the one that's easy to drop.
• Integrations can't be confirmed. The buyer's stack is a must-have, and a vague "integrates with your tools" doesn't pass.
• No security information. IT can't approve what it can't read.
• No customers like the buyer. Same industry, same size, same region.
• The AI answer is wrong or empty. An assistant that says you lack a feature, or doesn't mention you, has made the decision for the buyer.
• Hard to try. A demo request is a meeting; a trial is a test the buyer controls.

A shortlist test you can run this week

Run all three parts. Each takes a few hours.

Part 1: the AI answer audit.

1. Collect 10 questions your buyers really ask, in their words. Good places to look: first emails from inbound leads, the opening minutes of discovery calls, and Search Console queries.
2. Ask each question in four AI assistants your buyers use, in a fresh session, with no mention of your company.
3. For each answer, record whether you're named, where you appear in the list, whether what it says about you is right, which competitors appear, and which sources it cites.
4. Fix what you control: the pages it cites, the facts it gets wrong, the questions it can't answer from your site. Rerun in six weeks.

Part 2: shortlist questions in win/loss interviews. Ask won and lost buyers:

• Which vendors were on your first list, and where did you hear about each?
• What got a vendor cut before you spoke to anyone?
• What did you need to know that you couldn't find without contacting us?
• When did you first contact us, and what made it the right moment?

Part 3: the rep-free audit. Ask someone outside sales to answer your buyers' must-haves from your website alone: price, integrations, security, implementation time, contract terms and a similar customer. Time it and note each answer they couldn't find.

Question (buyer's words): [ ]
Assistant: [name]   Date: [ ]   Fresh session: Y/N

Are we named?  Y / N     Position in the list: [ ]
What it says about us: [ ]
Correct?  Y / partly / N     What's wrong: [ ]
Competitors named: [ ]
Sources it cites: [ ]
What we'll fix (page, fact or missing answer): [ ]

Worked example: an inspection software vendor runs the test

Illustrative example. The company, numbers and answers are invented to show the method.

The vendor: Quillon Inspect (quillon.example) sells property inspection software to residential property managers with 500 to 5,000 units.

Part 1: 10 questions across four assistants gave 40 answers. Quillon was named in 3; its two largest competitors in 31 and 26. Two of the three answers that named Quillon said it had no mobile offline mode, which it had shipped in May. The sources cited most often were review sites and competitors' comparison pages.

Part 2: of 12 lost deals interviewed, 7 said Quillon wasn't on their first list; 4 of those found it later through a peer. Of 9 wins, 6 started from a peer referral.

Part 3: a finance colleague couldn't find a price, an integration list or anything on security in 25 minutes.

What Quillon changed: it published price ranges by portfolio size, an integration page naming the property management systems it connects to, a security page, and a dated changelog entry for offline mode. Customer success asked happy customers for reviews. Six weeks later, Quillon was named in 11 of 40 answers and none repeated the offline-mode error. With samples this small, that's a direction, not proof; the team now reruns the audit each quarter.

Getting on the list is half of it. Once you're in the conversation, the buyer wants proof you understand their company; Bebop researches each account and writes a playbook with what's changed there and the angle to open with.

If you're the buyer: a shortlist worksheet

Decision: [what we're buying]   Owner: [name]   Decide by: [date]

MUST-HAVES (fail any one and the vendor is out)
1. [e.g. integrates with our CRM, two-way]
2. [e.g. SOC 2 Type II report available]
3. [e.g. total first-year cost under $X]

WEIGHTED CRITERIA (weights add to 100)
Fit to our use case: [30]   Total cost: [25]   Implementation: [20]
Vendor safety and terms: [15]   Ease of working together: [10]

VENDOR      Must-haves met?   Score   Evidence (source, date)
[A]
[B]
[C]

WHO HAS SIGNED OFF: user lead / budget owner / IT / procurement / finance

Questions worth asking every vendor on the list:

• What does a customer our size pay in year one, all in?
• Which of our must-haves do you not meet, or meet only partly?
• Can we test on our own data before we sign, and for how long?
• Can we speak to a customer in our industry who has been live for more than a year?
• What does implementation take from our side, in hours and people?
• What happens to our data if we leave?

Common mistakes

• Treating inbound leads as early. Most arrive with a favorite. Ask who else they're looking at and why.
• Gating the facts buyers use to cut vendors. Price, integrations and security pages are shortlist pages.
• Ignoring hidden buyers. Procurement, finance and IT shape the list; give them something to read.
• Never checking what AI assistants say. It takes an afternoon, and errors there cost deals quietly.
• Measuring only the deals you saw. Ask lost buyers whether you were ever on the first list.

Next step: once you're on the list, find out who can actually approve the purchase with the guide to who has buying authority.