The lead form arrives and everybody gets briefly excited. Somebody schedules the discovery call and somebody else pulls the account history. Then, forty minutes into the conversation, you realise the prospect has already read your pricing page, compared you against two competitors you weren't aware you had, and formed a firm preference about which of the three they'd rather work with. You aren't at the beginning of a sales cycle at all. You're auditioning for a decision that was substantially taken several weeks ago.

Sound familiar? Buyers moved, and sellers have been slow to adapt.

The evidence is unusually consistent

Three independent studies land in more or less the same place, which happens rarely enough in B2B research to be worth pausing on. According to 6sense research covering more than 4,000 buyers, 94% of buying groups had ranked their preferred vendors before making contact with any of them, and 77% eventually purchased from that preliminary favourite. Green Hat's APAC buyer study of 632 buyers reaches a similar conclusion from a different direction. Around 60% of the journey is complete before a buyer engages a vendor, most have settled their requirements by that point, and 76% say the first vendor they contact usually wins.

Read that last finding again.

The first vendor contacted usually wins, and the contact only happens once the shortlist already exists. So the commercial question in front of you isn't how to convert inbound enquiries more efficiently. It's how to become one of the two or three organisations a buyer can recall without prompting, in the moment they start assembling the list.

Which is a very different brief, isn't it?

The window is getting shorter, not longer

Here's the kicker, and it's the part that should change your strategy rather than your reporting. For years the received wisdom held that B2B buyers had gone dark and gone slow, disappearing into meandering, invisible research phases that marketing had to wait out patiently. But the current evidence suggests something less comfortable than that.

6sense's research-to-engagement split has moved from roughly 70/30 to 60/40, purchase cycles have compressed from 11.3 months to 10.1, and buyers now engage vendors six to seven weeks earlier. Dark and fast is a harder problem than dark and slow. A compressing research phase means a compressing window for influencing a preference, and the influencing still has to happen inside the invisible part of the journey, where you have no visibility and no invitation.

None of this is unfamiliar behaviour, incidentally. It's the same self-directed buyer marketers have been describing for a decade, simply operating at higher speed. A survey by Forrester Consulting found that 74% of B2B buyers prefer to research online rather than interact with a sales representative. The same independence shows up in Google's research, in McKinsey's work on remote and digital self-service, and in Amazon's reporting on self-serve purchasing. The direction of travel has been legible for years. What has changed is how little time it now leaves you.

They have already spoken to your customers

On an episode of the SaaS Stories podcast I had the pleasure of speaking with Rachael McBrearty, Chief Customer Officer at Evisort. We talked about what enterprise buyers are actually doing during that invisible stretch.

"They've usually done a lot of research, and they've probably already talked to some of your customers, sometimes before you even know it. They don't wait for you to give them a reference; they'll find one themselves."

Listen to the full episode of SaaS Stories with Rachael McBrearty

That observation reframes the reference customer entirely. A carefully curated case study handed over at proposal stage is a late-cycle asset, arriving well after the buyer has run their own back-channel and reached a private conclusion. Your genuine reference program isn't the PDF. It's whatever your customers happen to say in a Slack community, in a LinkedIn comment thread, or during a five-minute phone call you will never hear about.

Which raises an awkward question. Do you know what they're saying?

The search layer is moving as well

Discoverability used to be a reasonably well-understood problem. However, that understanding is shifting underneath everybody at once. Pew Research Center's behavioural panel, covering 68,879 real searches, found that users clicked a traditional search result in 8% of visits where an AI summary appeared, against 15% of visits with no summary present. Around 40% of companies are already investing in generative engine optimisation.

So the mechanics of being found are being rewritten while the window for being found contracts. Those two developments compound, and they compound hardest against the vendor with thin, undifferentiated, mid-funnel content.

What you can do about it is set out at the foot of this piece, in three moves.

Where this leaves you

The uncomfortable implication of the shortlist research is that a considerable proportion of demand-capture investment competes for buyers whose minds are made up. Demand creation is the harder, slower and less attributable half of the job. But it's the half that determines whose name gets written down first.

So here's a practical starting point. Pull your last ten closed-won deals alongside your last ten closed-lost. Then establish, by asking the buyers directly rather than inferring from the CRM, when each of them first encountered your organisation, and who else made their list. Most teams have never asked. The answers tend to reallocate a marketing budget faster than any attribution model manages to.

In the next piece I'll look at what happens once you do have the list of accounts you want, and why naming your targets turns out to be the straightforward part.