Measuring B2B social

This is the hardest measurement problem in the guide, because the channel's main effect is one your analytics cannot see.

Dark social is most of it

Someone reads a post, does not click, remembers you, and searches for you six weeks later. That conversion is recorded as branded search or direct. LinkedIn gets no credit and looks like it does nothing.

This is not a tracking bug to fix. It is how the channel works. Any measurement approach that ignores it will systematically recommend cutting the thing that is working.

What to actually measure

Self-reported attribution. A "how did you hear about us" field on the demo or signup form. Crude, biased, and the only instrument that sees dark social at all. Collect it as free text, read it monthly, and take it seriously even though it is unscientific.

Branded search volume. If LinkedIn is building awareness, more people search your name. This is a proxy, it moves earlier than pipeline, and it is hard to fake.

Pipeline influence. Tag opportunities where any contact engaged with LinkedIn content before the opportunity opened. Not causal, but far more honest than last-click.

Direct traffic. Blunt, but it moves when a channel is working and nothing else changed.

Experiments beat models

The strongest evidence available is a holdout: run the programme in some regions or account segments and not others, then compare pipeline. Expensive in patience, decisive in conclusion, and the only thing that will survive a sceptical finance conversation.

What to report

Do not hand a stakeholder impressions and follower growth. Report:

• What we published and to whom.
• Movement in the proxies — branded search, direct, self-reported mentions.
• Pipeline where any LinkedIn touch preceded it, stated as influence not credit.
• What we are testing next, and what result would change the plan.

Say plainly which numbers are proxies. Credibility is the asset; overstating attribution spends it.

What trips people up

• Reporting vanity metrics because they are available, then losing the budget when someone asks about revenue.
• Claiming attributed pipeline the CRM cannot support.
• Giving up after one quarter on a channel with a two-quarter lag.

You have got this when

You can defend the LinkedIn budget with three numbers, and honestly label which of them are proxies.