How B2B Revenue Attribution Improves Decisions

Self-reported attribution is inexpensive and adds a useful data point when someone recalls a distinctive marketing interaction. Dreamdata tested it across 100 booked demo calls, comparing answers with journeys recorded in its platform. Hedebrandt says people often forgot how they entered the funnel, while few responses offered useful detail.

Answers can also be too vague for investment decisions. “Google” does not separate advertisements from organic visits or identify campaigns and landing pages. Teams also risk confirmation bias by overstating a podcast’s impact after one mention. Hedebrandt recommends collecting self-reported attribution but not treating it as a definitive measurement system.

B2B journeys span time, sessions and people. Hedebrandt cites Dreamdata benchmarks showing an average 192-day path from first touch to a won account. These journeys involved more than 31 sessions and more than two people per deal. Tools centred on one visit, conversion or advertising click cannot readily represent that complexity.

Google Analytics lacks the full account journey, while a CRM often records only the final converting session. Advertising platforms show spending and clicks but not what happens months later. Dreamdata extracts available digital interactions, cleans and deduplicates them, then builds an account-based timeline. It cannot capture unrecorded offline conversations.

Dreamdata uses cookies and local storage, but Hedebrandt says disappearing cookies cause less damage in B2B than expected. Tracking resumes when a known user returns through an identifiable interaction, such as a newsletter link. The data warehouse retains earlier activity and adds each new visit, source and date. It does not require one uninterrupted browser cookie across every session.

Hedebrandt does not present attribution as providing 100 per cent visibility. Its purpose is better decision-making. Capturing 60 or 70 per cent of activity can distinguish weak marketing from productive marketing. Teams can then shift resources from unsuccessful work towards effective activity, improving their go-to-market approach over time.

Hedebrandt describes attribution as one use case within a product that gives B2B go-to-market teams a shared source of truth. At a previous company, advertising spending rose from zero to €100,000 monthly. Yet analysis compared each month’s spending with sales from that month. Customer journeys lasted six or 12 months, so clicks and recognised revenue did not align neatly.

His co-founders faced a related issue while responsible for Trustpilot.com’s product. They wanted to understand the period between account creation and a later sale. Connected account timelines reveal entry sources, journeys before churn or upsell, and activity across marketing, sales and customer success.

Companies should collect underlying data before they urgently need analysis because meaningful attribution requires historical evidence. Hedebrandt compares this timing to planting a tree: the best earlier moment has passed, making now the next best moment. Teams should not postpone tracking when measurement becomes necessary, though useful analysis takes time.

Pre-revenue companies with little website traffic should prioritise generating activity over sophisticated attribution. Early teams should execute, observe and ask where revenue originates. Stronger systems matter as journeys grow longer and involve more sessions. Hedebrandt also warns that many first-touch models show only the final session’s first touch, not the account journey’s true beginning.

Teams should begin by examining the paths taken by accounts won during the past year. They can repeat activities linked with wins and stop funding work that produced email addresses instead of pipeline or customers. Since money can be spent once, teams should direct it towards sales, marketing or customer success activity that produces the most revenue.

The evidence can also test departmental assumptions. Marketing can show how many won accounts it sourced, while conference data can expose leads that never entered the pipeline. Timelines reveal where strong trials began, which product features prospects used and which contacts preceded purchases. Teams can then assess whether each digital touch advanced accounts towards pipeline and won deals.

A smaller ABM dataset limits machine-learning analysis but retains operational value. An account executive handling 30 accounts can examine website visits, review-platform activity, LinkedIn advertisements and webinar participation. This account-level activity helps salespeople tailor outreach and conversations, despite having too few interactions for sound statistical conclusions.

Data availability remains the constraint. No platform can reconstruct journeys when businesses have few digital touchpoints, use unconnected calling software or leave offline meetings unrecorded. Hedebrandt distinguishes low volume from absent instrumentation. Limited account activity still offers useful sales intelligence. However, a B2B organisation without foundational systems such as a CRM is not an ideal fit.

Dreamdata made progress by defining an ideal customer profile through verifiable criteria, including geography, industry, employee count and CRM system. This discipline means declining poor-fit prospects. It also helps marketing avoid unsuitable campaigns, sales stop pursuing mismatched accounts and product teams reject irrelevant features. This focus matters when an organisation cannot saturate an entire market.

The company also learned to follow its team’s strengths. Investors encouraged outbound activity, but three to six months produced little progress. Dreamdata shifted to targeted advertising and inbound tactics matching Hedebrandt’s marketing experience. It reviews lead and account sources, expands productive activities and stops ineffective experiments. Less measurable work, including relevant podcast appearances, still requires judgement about audience fit and experience.