How do you measure B2B marketing?

The measurement argument usually starts before the dashboard

Most B2B measurement problems are presented as data problems. Quite a few are really agreement problems. Marketing thinks an influenced opportunity counts. Sales thinks sourced pipeline counts. Finance wants closed revenue and payback. The media team is optimising to cost per lead. The brand team is looking at search and awareness. Everybody can be technically correct and still be running five different versions of success.

Before adding another reporting layer, agree what the business is trying to move and which evidence matters at each stage.

Level 1: is the market responding?

These are the signals that let you steer while the commercial outcomes are still forming.

Depending on the programme, that can include reach into the right accounts, buying-group coverage, qualified site behaviour, branded search, direct traffic, content consumption, event response, repeat engagement, high-intent actions, sales acceptance and the quality of inbound conversations.

The word qualified matters. More activity is not automatically better. If engagement rises in audiences that never become customers, the metric has become a comfort blanket.

Level 2: is marketing changing pipeline?

This is where the conversation gets commercially useful: opportunity creation, pipeline sourced, pipeline influenced, stage progression, meeting quality, velocity, win rate and expansion inside the segments or accounts the strategy actually targeted.

Keep sourced and influenced separate. Combining them into one enormous number might make a slide look stronger, but it makes the methodology harder to trust. In complex B2B, influence is often the bigger story because several people and many touches shape the decision. That does not make every touch equally causal.

The measurement should help you decide what to do next, not only defend what you already spent.

Level 3: did the business outcome improve?

Revenue, gross margin where relevant, customer acquisition cost, payback, retention, expansion and efficiency belong here.

These outcomes move more slowly and are affected by more than marketing. That is not a reason to avoid them. It is a reason to be precise about contribution and timing.

Brand belongs here too, over a longer window. Stronger baseline demand, better quality inbound, more branded search, better conversion, greater direct traffic, improved win rates and pricing resilience can all matter. Judging brand purely by leads generated this month guarantees the channels nearest the form will keep winning the budget conversation.

Demand captured and demand created are not the same job

Search, review sites, direct traffic and high-intent pages often capture demand that already exists. They are valuable because somebody still has to win that demand. But they should not get sole credit for creating the desire that led the buyer there.

Brand, content, events, communities, programmatic, social, partners, PR and sales activity can all shape demand before it becomes visible. Some of that influence will never be perfectly tracked. A colleague forwards a deck. Someone asks an AI assistant for a shortlist. A buyer sees three posts but never clicks. A procurement lead hears your name from a peer. Then another person types the brand into Google and the final touch gets crowned.

Separate capture from creation in the way you interpret the numbers. Otherwise the business keeps investing in harvesting the existing pool and wonders why the pool is not growing.

Not every channel should be judged by the same KPI

If every channel is optimised to cost per lead, the plan slowly becomes a lead-capture machine. Channels designed to create familiarity, reach buying groups or influence a shortlist get punished because their effect appears elsewhere.

Give each part of the plan a job. Search might capture active demand. Programmatic might build account coverage and frequency. LinkedIn might distribute points of view and reach defined roles. Events might create access and deepen relationships. Content might help both discovery and sales conversations. Then connect those jobs to the commercial ladder above.

This is more useful than pretending a podcast impression and a pricing-page visit should compete on the same metric.

Attribution is a lens. Use more than one

There is no attribution model that reconstructs a B2B buying decision exactly. Last click over-rewards what happened nearest the conversion. First touch invents certainty at the other end. Multi-touch models divide credit according to rules the business chose. Contact-level tracking misses the other people in the account.

Use a consistent model so trends are comparable, but triangulate it with self-reported attribution, account-level engagement, sales feedback, CRM movement and incrementality tests where spend and scale make them possible. When the evidence disagrees, investigate the disagreement instead of forcing the numbers to reconcile.

A measurement model becomes more credible when it is willing to say we do not know exactly and still shows enough evidence to make a sensible investment decision.

What the board actually needs

Most boards do not need the channel dashboard. They need a stable commercial view: pipeline movement in the markets that matter, revenue contribution, efficiency, major changes in demand and the few leading indicators that have proved useful.

Keep the format consistent in good and bad quarters. If the definition of success changes every time the number disappoints, trust disappears faster than budget.