Aggregate the signals from every contact in the account into one account score, because the buying committee is buying together and scoring leads in isolation misses the picture that matters
Most B2B lead scoring is contact-level, which means it treats each lead as if they were the unit of decision and misses the structure of how the account actually buys. A B2B purchase is a buying committee with multiple people from the same company looking at the product, each with different roles, each leaving different signals, and the decision to engage commercially is made by the committee, not by any one contact. A score that says "this contact is highly engaged" can be a champion deep in evaluation, or it can be an analyst doing research while the actual buyers have not looked at the product yet. The two situations need different next actions, and the contact-level score cannot distinguish them.
Account-level scoring rolls every contact's signals up to the account and weights them by role, so the score reflects the account's collective engagement rather than the most-engaged individual's. Three contacts engaging lightly across different departments is a stronger account signal than one contact engaging deeply alone, and the routing the score drives needs to reflect that. The recipe is medium-high readiness because the data discipline is real and the score has to mean something the sales team can act on, but the building blocks are well-understood once the account resolution is in place.
This recipe builds the score at the account level with role-weighted aggregation, and uses it to drive routing the sales team trusts.
An account-level engagement score that captures the buying committee's collective signal, with the role-weighting and recency discipline that makes it actionable. The metric is the score's correlation with pipeline conversion: accounts scoring above the routing threshold should convert to qualified pipeline at a materially higher rate than accounts below it, measured per quarter and per segment. The value shows up in routing that delivers accounts ready for sales conversation rather than contacts that look engaged in isolation, and in the BDR's working day (the queue is accounts the committee is engaging, not contacts engaging while their committee is silent).
The second outcome is the visibility of the buying committee itself: the score makes the multi-contact pattern visible to the team, which surfaces accounts where one contact is engaging deeply and the committee around them is silent (a signal worth treating differently from broad-committee engagement).
A warehouse-side aggregation that rolls contact events to the account, with the weighting logic, the recency decay, and the threshold that drives routing. Composable stacks fit because the events arrive at the warehouse and the score is a query against them. The capability that decides whether the score is actionable is the role-weighting calibration: a flat aggregation that treats every contact's signal equally produces a score that the analytics intern's research-tier engagement contributes the same as the VP's evaluation engagement, and the routing built on it pulls the team toward accounts where the loudest signals come from people who do not decide.
Compare the tools on Martech Stack Builder
Marketing ops owns the score definition, the contact-role weights, the recency decay, and the threshold for routing. Data science is involved where the aggregation moves beyond rules to a model (logistic regression on historical pipeline outcomes against the engagement features), though most teams start with a transparent rule-based score and graduate as they have outcome data. Sales ops owns the routing the score drives and the qualification standards the score has to meet, because a score that produces accounts sales does not pursue is a score with the wrong threshold or the wrong weights. Data engineering owns the contact-to-account roll-up pipeline and the freshness discipline that keeps the score current. The recipe takes months because the score's quality is a function of the engagement data and the contact-to-account mapping, and both take time to clean up. A score deployed against partial events or stale mappings produces routing that sales loses faith in, and rebuilding the faith is more work than getting the data right the first time.
Work in this order. Calibration comes before routing, or sales chases noise.
Household or account-level resolution covers steps 1 and 2, intent signal aggregation steps 3 and 4, and behavioral segmentation steps 5 and 6.
The first failure is the role-blind aggregation. A score that adds engagement events without weighting them by who produced them rewards activity patterns that look strong from analyst research but predict nothing about commercial readiness, and sales pursues accounts whose committees are silent because the analyst engagement is loud. Role weights, calibrated against historical pipeline outcomes, are what make the score correlate with what the team actually cares about.
The second is the stale contact-to-account mapping. A contact who moved to a different company three months ago is still rolling engagement to the old account if the mapping has not refreshed, which contaminates the score in two directions: the old account looks more engaged than it is and the new account looks less. A reconciliation cadence (against the CRM, against LinkedIn, against email-bounce signals) is what keeps the mapping current, and a recipe without it watches the score quality erode as the corporate mobility of the audience does its work.
The third is the score that does not differentiate from the contact-level score it replaces. If sales sees the account score change every time a contact opens an email, the routing collapses into the same noise as the contact-level scoring, and the buying-committee picture is lost in the volatility. The recency decay and the threshold work together: the score smooths over individual events and shifts on patterns, so the routing fires on accounts genuinely entering engagement rather than on accounts where one contact happened to scroll a blog post.
The Workshop works out with your team which of these matter for your stack right now, and what to do first: a 90-minute session with the people who own the decision.
The contact-to-account roll-up that captures the committee, the role-weighting that distinguishes the VP from the analyst, the recency discipline that holds the score steady, and the threshold that produces routing the sales team trusts: those are the decisions that turn lead scoring into account scoring the buying motion actually responds to.
Did this recipe match your situation?Anonymous response. Sign up to leave a longer note tied to your account.
Activate the target account list with one identity resolution and one exclusion list across paid display, paid social, direct mail, and BDR outreach, so the same account stops getting contradictory treatment
Lower acquisition cost by feeding the bidding algorithms better conversions, deduplicated, consent-valid, and weighted by value, rather than just more of them
Target CTV from your own customer segments and measure the reach you actually added above linear, rather than buying impressions on faith