Lead transitions from marketing to sales with the full intent context attached, account-aware rather than lead-isolated
The standard B2B handoff is a row in the CRM with a name, an email, a company, and a lead source. Sales calls cold or close to it. Marketing has been watching this lead read four whitepapers, attend a webinar, download a comparison guide, and visit the pricing page twice. None of that context travels with the handoff. Sales asks questions marketing already has answers to. The prospect loses patience. The cycle takes longer than it should and the conversion rate is lower than it should be.
The other failure mode is timing. Marketing passes leads too early, when the prospect was still researching for someone else's project. Sales rejects the leads, refuses to follow up, and the relationship between the two functions degrades. Marketing then over-corrects and passes leads too late, after the prospect has already evaluated three competitors and is in pricing negotiations elsewhere. Both failures stem from the same underlying problem: the handoff threshold and the handoff payload are typically defined independently of each other, and neither is calibrated against the signals that actually predict conversion.
The recipe addresses this by aggregating intent signals across lead and account levels, scoring them with velocity awareness rather than simple accumulation, triggering enrichment at the threshold crossing rather than at every lead capture, and assembling a handoff payload that includes the full activity history so sales has context from the first interaction.
A lead transitions from marketing to sales when the intent signals (weighted by recency, by topic relevance, and by signal type) cross a defined threshold. The transition includes account-level context: what this lead has done, alongside what others from the same company have done in parallel. The handoff payload includes the activity history (pages visited, content downloaded, demo requests, email engagement, product activity if PLG), the firmographic context (company size, industry, technology stack signals), the ICP fit score, and a recommended next action. The sales tool surfaces this context at first contact, so the call or email opens from where the prospect actually is rather than from zero.
A marketing automation platform (Marketo, HubSpot, Pardot, or similar) for lead capture, email engagement tracking, and the workflow that triggers the handoff. A CRM (Salesforce, HubSpot CRM, or similar) as the destination for the handoff payload. A lead scoring or intent scoring system, either built into the marketing automation platform, built into a CDP, or sourced from a dedicated intent provider (6sense, Demandbase, Bombora). An enrichment service appended at the threshold event. A CDP, warehouse, or RevOps data layer for account-level aggregation, especially if the marketing automation platform's account-aware scoring is too limited. A sales engagement tool (Outreach, Salesloft, Apollo) that can consume the handoff payload and surface the context to the sales rep.
Compare the tools on Martech Stack Builder
Five roles. Sales operations owns the threshold and the relationship with sales: what counts as a handoff, what territory rules apply, what the sales rep sees when the lead lands, and the feedback loop from sales disposition back into the scoring model. Marketing operations owns the signal weighting and the content-to-intent mapping: which whitepaper indicates which stage, which page visit weights more than another, what counts as a strong versus a weak signal. Data engineering builds the score calculation infrastructure and the account-level aggregation logic, especially the velocity calculation that turns signal accumulation into a more predictive measure. Identity or CDP ownership manages the lead-to-account binding, ensuring that signals from multiple leads at the same company aggregate to the right account rather than scattering across orphaned records. Analytics measures conversion rates across the handoff cohorts, calibrates the threshold against actual outcomes, and provides the evidence that justifies weight adjustments.
The intent signal score has three components and two scaling dimensions. The components are signal type (page views, content downloads, email engagement, demo requests, product activity, account-level peer signals), signal weight (how much each signal type contributes), and signal recency (how fast the contribution decays over time). The scaling dimensions are lead-level (this individual's behavior) and account-level (cumulative behavior from all known contacts at this company).
Signal weighting is the work that most teams underinvest in. The default weights from the marketing automation platform are a starting point, not an answer. They reflect averages across many customers, not the patterns specific to this company's funnel. The calibration process is to take historical conversion data, identify which signals preceded conversion at meaningful rates, and weight them accordingly. The first calibration is rough; subsequent rounds refine it as more conversion data accumulates. Mature scoring models update weights quarterly with explicit governance.
Velocity matters more than accumulation. A lead that picks up two hundred points over six months is in slow research mode; the lead might convert eventually but isn't ready for a sales call this week. A lead that picks up one hundred and fifty points in three days is showing intent acceleration; this lead is in market now. Threshold logic that triggers on point totals over-fires on the slow researchers and under-fires on the fast accelerators. Threshold logic that triggers on point velocity (rate of accumulation) catches the fast accelerators correctly. Most modern scoring models include both: a baseline point threshold plus a velocity multiplier.
Account-level aggregation is what makes the recipe B2B-appropriate rather than B2C-with-firmographics. One lead reading three whitepapers is a lead. Five leads from the same domain, all reading whitepapers in the same week, is an account showing buying-committee activity. The account-level signal is structurally stronger than the lead-level signal because it indicates organizational interest rather than individual research. The aggregation has to be identity-aware: the same person reading the same whitepaper twice shouldn't double-count, but five different people reading five different things should aggregate appropriately. The identifier graph carries this work.
Enrichment at threshold is a cost optimization. Enrichment APIs charge per lookup. Calling them on every form submission produces a high volume of enrichment events for leads that will never convert. Calling them at the threshold crossing produces enrichment events only for leads about to be handed off, where the enrichment is actually useful. The enriched data (company size, industry, technologies, recent news) gets attached to the handoff payload, giving sales the firmographic context they need without the per-lead enrichment cost.
The handoff payload itself is structured for fast sales consumption. The first item is the recommended next action (call this week, send a personalized email, schedule a demo through self-serve, wait and watch). The second is the intent context summary in a sentence or two: what topics they've engaged, what stage they appear to be in, what triggered the threshold crossing. The third is the activity timeline, scannable in fifteen seconds. The fourth is the account context, including other contacts at the same company and their activity. The sales rep should be able to open the lead and have a useful first conversation within sixty seconds of seeing the record.
The threshold calibration trap catches almost every implementation. Marketing sets the threshold at one hundred and fifty points based on a feel for "what looks like a real lead." Sales accepts the first month of handoffs, then starts rejecting them. The team blames marketing for sending bad leads. The actual problem is that one hundred and fifty points is the wrong threshold for this specific business at this specific time. The fix is to calibrate from data: look at leads that converted historically, find what their score was at the moment they actually engaged with sales successfully, and use that as the baseline. The first calibration will be wrong. Run it for a quarter and adjust.
The signal-burst over-fire is the velocity trap from the other direction. A prospect publishes a blog post on their company website that gets shared internally. Fifteen people from one company hit the brand's site in three days reading the same three pages. Account-level score spikes. The threshold crosses. Sales gets a handoff that doesn't actually correspond to buying intent; it corresponds to a content-driven research burst. Velocity calculation needs a smoothing factor and ideally a signal-type filter that distinguishes research bursts from buying bursts.
The "marketing isn't sending good leads" feedback loop is the political failure mode. Sales rejects handoffs. Marketing adjusts the threshold higher to reduce volume. Sales accepts a smaller number but still rejects most of them. Marketing assumes sales just doesn't want to work the leads. Sales assumes marketing doesn't understand what a good lead looks like. The fix is structural: build a feedback loop where sales disposition (accepted, rejected, converted, lost) flows back into the scoring model as training data, with regular reviews of the disposition reasons. The conversation moves from "your leads are bad" to "the model needs to learn from these specific disposition signals."
The orphaned-account problem is the identity failure that breaks account-level aggregation. Two leads come in from the same company over six months, but they used slightly different company name formats ("Acme Inc" and "Acme, Inc."), and the system created two separate account records. The account-level aggregation runs against each, missing the cross-record signal that would have justified a handoff. The fix is account-level identity resolution with deterministic merge rules based on email domain, normalized company name, and corporate hierarchy data from enrichment.
The intent-data illusion catches teams that buy third-party intent signals (Bombora, 6sense, third-party providers) and treat them as direct evidence of buying intent. These signals are useful as input to a scoring model but are not, on their own, reliable triggers. The data is generated from inferred behavior across publisher networks, with confidence levels that vary widely. Using it as a threshold trigger without weighting against first-party signals produces false positives at a rate that erodes sales' trust in the handoff. Treat third-party intent as a feature, not a verdict.
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.
It's a missing calibration discipline, an account-level identity gap, or a scoring model that hasn't been updated since it was first configured. The Martech Workshop is where we work out which of those is true for your stack, and what the path to a calibrated, account-aware, velocity-respecting handoff looks like from where you currently are.
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
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
Lower acquisition cost by feeding the bidding algorithms better conversions, deduplicated, consent-valid, and weighted by value, rather than just more of them