Short answer. The six signals that predict pipeline from event leads are: budget mention, timeline mention, named pain, named competitor, decision-maker role, and explicit next-step ask. Default lead scoring models capture none of them because none of them appear on a form fill. We use Confee as the working example for capture because these signals only show up in the conversation, not on a badge scan.
This is for RevOps and Demand Gen teams setting up scoring for conference and trade show leads in HubSpot, Marketo, or Pardot.
Key takeaways
- Default lead scoring models reward form completeness, not conversation depth. Event leads have neither.
- The signals that matter come from the conversation: budget, timeline, pain, competitor, role, and next step.
- A lead with a budget mention is roughly 5x more likely to close than one without.
- Confee captures all six signals automatically as structured fields, ready to feed your scoring model.
Why default scoring fails for event leads
Most scoring models are built for inbound forms. They reward:
- Filled-in fields (email, company, role).
- Page views before the form (engagement).
- Lead source (paid vs organic).
Event leads have one filled-in field (the badge scan, often just an email) and zero page views before the booth. Plug them into a default model and they score low, even when the conversation was strong.
The fix is to score on conversation signals, not form completeness.
The six signals that predict pipeline
Based on close-rate analysis across B2B SaaS event programs, six signals stand out.
Signal 1: Budget mention
When the prospect names a number, even a rough one, close rate jumps roughly 5x.
- Examples: "We have around €50k earmarked." "Our budget is in the low six figures."
- Score weight: high (15-25 points).
Signal 2: Timeline mention
When the prospect names a window for a decision, close rate jumps another 2-3x.
- Examples: "End of Q2." "Before our renewal in October." "Right after our board meeting."
- Score weight: high (15-20 points).
Signal 3: Named pain
A specific, named pain point in their words. Generic complaints score lower than specific ones.
- Strong: "We lose leads after every event because reps forget to log them."
- Weak: "Sales is tough." "Our CRM is messy."
- Score weight: medium-high (10-15 points).
Signal 4: Named competitor
When the prospect mentions they are evaluating a competitor, they are actively in market.
- Examples: "We've looked at Gong." "We use Otter for some calls."
- Score weight: medium (10 points).
Signal 5: Decision-maker role
VP, Director, Head, or C-level signals direct buying authority. Manager and IC roles signal influence, not authority.
- Score weight: medium (5-15 points based on level).
Signal 6: Explicit next-step ask
The prospect asks for a follow-up unprompted.
- Examples: "Can you send me the deck?" "I want to set up a call with my team."
- Score weight: high (15-20 points). The strongest single signal.
Sample scoring model
A workable starting point. Tune from your own close-rate data after one quarter.
Budget mention +20
Timeline mention +15
Named pain (specific) +12
Named pain (generic) +5
Named competitor +10
Decision-maker (C-level) +15
Decision-maker (VP/Director) +10
Decision-maker (Manager) +5
Explicit next-step ask +18
Threshold: 40+ for AE assignment. Under 40 routes to nurture.
Total points possible: roughly 100. A lead that mentions budget, timeline, and asks for a demo is already over the threshold even without role data.
How Confee captures these signals
The hard part is not designing the scoring model. It is capturing the inputs.
Confee extracts each of these fields automatically from the booth conversation:
- Budget. Pulls numbers in currency context.
- Timeline. Pulls dates, quarters, and time references.
- Pain. Captures the prospect's own phrasing of their problem.
- Competitor. Flags any mention of a competing product or vendor.
- Role. Extracts from "I'm the VP of X at Y."
- Next-step ask. Detects explicit follow-up requests.
These fields land in the CRM as structured properties. Your scoring model reads them and assigns points automatically. No rep types any of it. For the routing layer that consumes these scores, see routing booth leads to AEs.
Common mistakes
- Scoring on attendance alone. "Visited Web Summit booth" is +5 points everywhere. That is fine as a base, but it cannot be the whole model.
- Ignoring the conversation. If you cannot score on conversation signals, you are scoring random noise.
- Over-weighting role. A C-level title without budget or timeline mention is just a name. Conversation signals beat title every time.
- Static models. Re-tune scoring weights every quarter based on actual close-rate data.
FAQ
Why does default lead scoring fail for events? Because it is built around form data and page views, neither of which exist for booth leads. Event leads need conversation-based scoring.
Which signal matters most? The explicit next-step ask. When a prospect proactively asks for a demo or a follow-up call, they have already pre-qualified themselves.
How do I get conversation signals into HubSpot or Salesforce? You need a capture tool that produces structured fields. Confee does this automatically. Manual logging works in theory but breaks in practice.
Can I score event leads in HubSpot's native scoring? Yes, once the conversation signals are in the CRM as properties. Native HubSpot scoring handles the rest.
How does Confee fit into a Marketo or Pardot stack? Confee syncs the structured fields into your CRM (Salesforce). Marketo and Pardot read those fields and apply your scoring model. Same flow as any other lead source.
Sources
- Confee CF-01 product documentation
- Forrester research on B2B lead scoring