Dark social. Where most B2B buying actually happens. And the demand gen approach that fits.
You will never get last-touch attribution to tell you the truth about how B2B buyers actually decide. They DM each other in Slack. They forward an article in WhatsApp. They mention a tool on a podcast. They ask a friend over coffee. None of that shows up in your dashboard. All of it shapes the decision. Here is the demand gen approach built around the reality.
What dark social actually is
Defined precisely.
Dark social is any digital sharing or referral that happens outside of the analytics infrastructure you control. Direct messages, private Slack threads, WhatsApp groups, voice notes, email forwards, pasted links without UTM parameters, screenshots of your content shared in private channels.
It also includes the audio referrals: the podcast where someone mentions you, the conference talk where your name comes up, the founder dinner where someone says "you should look at X". None of that is trackable. All of it influences buying.
B2B research consistently shows that 70-80% of B2B buying decisions are heavily influenced by these channels. Last-touch attribution captures the 20% that happens to click through with a UTM. Most of the work is invisible to your CRM.
How to do demand gen in a dark-social world
Six adjustments to a standard demand gen playbook.
Optimise for shareability, not for clicks
Posts that get DMed and forwarded matter more than posts that get clicks. Hooks, takeaways, framework names that travel. "I sent this to a friend" is the metric.
Author identity matters more than logo
Personal brand on LinkedIn outperforms company brand for share velocity. People DM each other someone's post, rarely a company's ad.
Build content for the screenshot
A single chart, framework, or quote that someone can screenshot and paste into Slack is more leveraged than the article it lived in.
Show up in the rooms
Podcast appearances, niche Slack communities, founder dinners, conference panels. Brand presence in places that do not show up in your analytics.
Self-reported attribution
Add "How did you hear about us?" to your demo form. The free-text answers tell you more about your real channels than any tracking dashboard.
Long content windows
Buyers come back 3, 6, 12 months after first hearing about you. Stop reporting 30-day attribution. Run 6-month and 12-month windows alongside.
Common mistakes
Five.
Cutting demand gen budgets because attribution is fuzzy. The activity that drives the most pipeline is the activity that is hardest to track. Cutting it because it is unmeasurable is a slow-motion own goal.
Doubling down on gated content to get more "tracked" leads. The leads are worse, the trust signal is worse, and the buyers who would have come through dark social are now annoyed by your forms.
Believing your CRM. The CRM tells you the last-touch source. It does not tell you why the buyer was on your site in the first place. Attribution is a hint, not a fact.
Killing podcast and event spend because the spreadsheet says they do not work. The spreadsheet is wrong. Both are dark-social rocket fuel. Track the right way: self-reported source on the demo form.
Trying to retrofit attribution onto dark channels with surveillance tooling. It does not work, it irritates buyers, and the data quality is worse than self-reported "where did you first hear about us".
How we run this for clients
In two sentences.
We bias the work towards founder-led content, podcast appearances, community presence, and shareable artefacts (frameworks, charts, original research). We measure with self-reported attribution on the demo form, supplemented by 6-month assisted attribution where the CRM allows.
FAQ
Self-reported attribution. Add "How did you hear about us?" to your demo form, run it for 90 days, show the answers to the CFO. Most companies discover that 30-50% of demos cite a source the dashboard never tracked. That is the justification.
Yes, but as part of a mix. Paid is cleaner to attribute and works for bottom-funnel intent capture. Dark-social channels work for top-funnel awareness. Run both. The mistake is running only what you can measure.
Always use them where possible. They will not solve the dark-social blind spot but they sharpen the parts you can see.
Yes, especially founder appearances on niche industry podcasts. The conversion rate is invisible but the brand recall is real. Compare it to other awareness channels, not to last-touch lead gen.
Want to talk through your situation?
Thirty minutes with the team. We will pressure test your numbers, not pitch.
Book a strategy call →