Updated July 2026. By Remy Beaumont, co-founder, Ignita.
Dark social B2B is where your deal is actually decided: the private Slack and WhatsApp groups, DMs, communities and podcasts where buyers research vendors long before they fill in a form. None of it passes a tracking parameter. All of it shapes the shortlist. This post lays out the evidence, why younger buyers research where you cannot see, and what to change about your launch strategy as a result.
Key takeaways
Millennials and Gen Z now make up 71% of B2B buyers, according to Forrester's Buyers' Journey Survey, and they research in private channels by default.
Buyers are roughly 70% of the way through their journey before they contact a vendor, and they initiate first contact around 80% of the time, per 6sense's Buyer Experience Report.
A Refine Labs study found a 90% gap between what attribution software claims and what customers self-report. Software credited web search with 78% of customers; the customers themselves credited it with 12%.
Only 17% of B2B buying time is spent with sales reps, and 61% of buyers now prefer a rep-free experience, according to Gartner.
The practical conclusion: your launch is not judged on your website. It is judged in Slack groups, DMs and group chats you will never see in a dashboard.
Why does this matter now?
Because the buyer changed and most B2B measurement did not. The people evaluating your product today grew up recommending restaurants in group chats and settling arguments with a screenshot. When they move into a buying committee, they carry that behaviour with them. Forrester found that millennials and Gen Z rose from 64% of B2B buyers in 2022 to 71% a year later, and they make up two thirds of buyers even on deals above $1m.
If your launch strategy is built purely on what your analytics can see, you are optimising the minority of the journey and ignoring the part where the decision is made. We covered who the new buyer is in Gen Z B2B buyers now run the deal. This post covers where they decide.
What does the research show?
Method note: this piece is a literature review of named third party studies, cross checked against the original publishers. The short answer: the visible funnel is a small window on a mostly private process. Five separate research bodies point the same way.
The journey is mostly seller-free. Gartner puts time spent with sales reps at 17% of the total buying journey, against 27% spent researching independently online. In 2025 Gartner reported that 61% of buyers prefer a rep-free experience altogether.
The decision forms before you know the buyer exists. 6sense found buyers are about 70% through their journey at first vendor contact, initiate that contact around 80% of the time, and 81% already have a preferred vendor when they reach out. In roughly 8 out of 10 journeys, that early favourite wins.
Your attribution is measuring the wrong thing. Refine Labs compared software attribution against self-reported attribution across 620 conversions and $21.5m in closed-won revenue. Software said 78% of customers came from web search; customers said 12%. Podcasts drove 53% of self-reported journeys and effectively 0% in software.
Buyers want to self-serve. TrustRadius found virtually 100% of buyers want to self-serve at least part of the journey, and Gen Z buyers are 35% more likely than any other generation to want it fully self-service.
Most of your market is not even shopping yet. The LinkedIn B2B Institute and Ehrenberg-Bass 95:5 rule holds that only about 5% of your category is in-market at any time. The other 95% is forming memory in exactly the channels attribution cannot see.
What is dark social in B2B?
Dark social is any sharing or discussion that analytics cannot attribute: links pasted into DMs, Slack and Teams threads, WhatsApp groups, community forums, podcast mentions, event conversations and screenshots. The term was coined by Alexis Madrigal in The Atlantic in 2012, when he noticed the majority of "direct" traffic was actually shared links with the referrer stripped. In B2B the stakes are higher than in consumer, because the unit of decision is a committee, and committees deliberate privately. When a head of growth asks "has anyone actually used this?" in a 400-person operator Slack, that thread does more for or against your brand than your entire paid budget that month, and it will show up in your dashboard, if at all, as "direct".
Why do younger buyers research where you cannot see?
Three reasons, all rational from the buyer's side. First, trust: younger buyers treat vendor content as claims and peer word as evidence, which is why TrustRadius finds them far more likely to lean on reviews and customer-led content. Second, efficiency: asking a community compresses weeks of evaluation into one thread. Third, safety: as we argued in the rational B2B buyer myth, B2B decisions are career decisions, and a peer endorsement in a private channel de-risks the choice in a way no case study can. The behaviour also extends to machines: many younger buyers now ask AI assistants to build the first shortlist, and both the community thread and the AI answer draw on the same raw material: what people say about you when you are not in the room.
Who is doing this well?
Three named examples worth studying. Gong built its category position on founder-led and employee-led LinkedIn content years before most sales-tech rivals took organic social seriously, so buyers arrived at sales conversations already convinced. Refine Labs grew almost entirely through its founder's LinkedIn presence and podcast, then published the attribution study above showing exactly why that engine never appeared in software. Notion's early growth ran through community: ambassadors, template creators and user groups doing the persuading in channels Notion did not control. None of these are tracked channels. All of them are deliberate strategies.
The Ignita insight: dark social is a launch strategy problem, not a measurement problem
Here is the conclusion most teams miss. The instinct, when the data goes dark, is to buy better attribution. But you cannot instrument a group chat, and you should stop trying. The correct response is to change what you launch, not how you track it. A launch optimised for the visible funnel produces an announcement. A launch optimised for dark social B2B produces an artefact worth pasting into a Slack thread: a benchmark, a teardown, a dataset, a demo clip, a founder post with an actual opinion. The test we apply to every client launch at Ignita is one question: would a stranger screenshot this and send it to a colleague? If the answer is no, the launch will perform exactly as well as the trackable channels allow, which the data above says is a small fraction of the journey.
This is also why cultural brand building compounds: the 95% who are not in-market today will not remember your feature list, but they will remember the take, the data or the founder voice that kept appearing in their feeds and group chats. That is the memory that surfaces when the buying trigger finally arrives. It is the thinking behind our guide to what a launch narrative is and why your startup needs one.
What founders and marketers should do
Add one attribution question everywhere. Put "how did you hear about us?" as a free-text field on every demo and sign-up form, and log answers verbatim. This is the cheapest window into dark social B2B you will ever build.
Launch with a shareable artefact, not just an announcement. Ship a benchmark, original dataset or teardown alongside the news. Give people something to paste into the thread where your category is being discussed.
Make the founder the channel. Buyers follow people, not logos. A founder posting specific, opinionated content weekly outperforms a brand account posting polished content daily.
Be present in communities before you need them. Join the 5 to 10 Slack, Discord and WhatsApp groups where your ICP actually talks. Contribute for months before you launch, so the launch lands with people who already know you.
Feed the answers buyers will find. Reviews, comparison threads and community mentions are what both peers and AI assistants cite. Actively ask happy customers to review and to talk.
Judge channels on pipeline conversations, not clicks. If self-reported answers keep naming a podcast or a community that software credits with nothing, fund it anyway. The Refine Labs data says the software is the one that is wrong.
If you are planning a launch and want it built for the channels that actually decide, talk to us about your launch narrative, or browse more research in our insights hub.
FAQ
What is dark social in B2B marketing?
Dark social is buyer research and sharing that analytics cannot attribute: DMs, private Slack and WhatsApp groups, communities, podcasts and word of mouth. It usually appears in dashboards as direct traffic.
Why is it called dark social?
The term comes from Alexis Madrigal's 2012 piece in The Atlantic, describing social sharing that arrives with no referrer data, making it invisible or "dark" to analytics tools.
How much of the B2B buying journey happens before sellers get involved?
6sense research puts first vendor contact at roughly 70% of the way through the journey, with buyers initiating contact around 80% of the time and 81% already holding a preferred vendor.
How do Gen Z and millennial B2B buyers research vendors?
Largely through self-serve and peer channels: reviews, communities, social content and AI assistants. TrustRadius found Gen Z buyers are 35% more likely than any other generation to want a fully self-service journey.
Can you measure dark social?
Not directly. The practical proxy is self-reported attribution: a free-text "how did you hear about us?" field, read verbatim. Refine Labs found a 90% gap between software attribution and what customers actually report.
Does dark social replace SEO and paid?
No. Search and paid still capture demand from in-market buyers. Dark social is where demand is created among the roughly 95% who are not yet in-market, per Ehrenberg-Bass and the LinkedIn B2B Institute.
What should a startup do at launch to win dark social?
Ship something worth sharing privately: original data, a benchmark or a strong founder take. Seed it in communities you joined before launch, and measure it through self-reported attribution rather than clicks.
We publish research like this every week. Join Ignita's free Substack here.
