---
name: webinar-attendee-scoring
description: Scores webinar and virtual event attendees for sales follow-up from the platform's attendee and engagement report (watch time, live or on-demand, poll answers, questions asked, resource and CTA clicks, repeat attendance) plus ICP fit and CRM status. Assigns lead tier A, B or C with reasons, turns each question into a one-line sales note and writes a talk track per lead. Use after a webinar or virtual event when someone has an attendee report and asks who sales should call first, or wants engagement scoring and a sales handoff.
license: MIT
metadata:
  author: eventmarketingsoftware.com
  version: "1.0"
  category: webinars-and-virtual
  source: https://eventmarketingsoftware.com/skills/webinar-attendee-scoring
---

# Webinar attendee scoring

Webinar platforms export rich engagement data that rarely reaches sales in a usable form. This skill turns the attendee report into a ranked handoff list: lead tier A, B or C, plain reasons, a one-line note from each person's questions and poll answers, and a talk track for the first call.

Lead tier, flags and campaign member status follow [references/shared-terms.md](references/shared-terms.md). The fit part of the score is the same rubric event-lead-scoring uses, so webinar and booth leads are comparable. For registrants who didn't watch at all, use no-show-recovery.

## Intake

1. Read `event-context.md` (or `event-program-context.md`) if present: ICP, target accounts, owners, CRM, tier rules.
2. Ask for anything missing, in one message:
   - The attendee report (and the separate engagement, poll, Q&A or chat exports if the platform splits them).
   - Session length in minutes, and the webinar topic.
   - Which poll answers mean "buying now" (for example "Evaluating now" or "In the next 3 months") and which CTA means "talk to sales".
   - ICP (industries, company size, departments, seniority). Defaults are in the rubric if none.
   - Owners and routing: CRM connector, an account owner export, or a round-robin list.
   - Webinar date (for due dates).

## Connectors (optional)

- **HubSpot or Salesforce connector**: read account owner, customer status and open opportunities by email and domain. Create tasks or update fields only after the approval gate.
- **Otherwise**: work from uploaded exports and return a CSV and Markdown tables.

## Method

### 1. Prepare the data
- Join the exports on email (lowercased). One row per person.
- Map columns: live minutes, replay minutes, poll answers (one column per poll), questions (text), chat messages, resource clicks, CTA clicks, past webinars attended.
- Set `attendance_type`: live, on-demand, or both. People with zero minutes everywhere are no-shows: move them out and point to no-show-recovery.
- Exclude competitors, internal staff, speakers and students with `lead_tier = Exclude` and a reason (flags as in shared terms).
- Count rows in, people out, joins that failed.

### 2. Score engagement (0 to 50)
Use [references/scoring-rubric.md](references/scoring-rubric.md):
- Watch share (minutes / session length, live or replay, whichever is higher).
- Buying-stage poll answers.
- Questions: product, pricing, integration or implementation questions score higher than general ones.
- CTA clicks: "talk to sales" or "book a demo" is the strongest signal; resource downloads are lighter.
- Repeat attendance across webinars.

### 3. Score fit (0 to 50)
Same rubric as event-lead-scoring: seniority, department, company size, industry, target account. If size and industry are missing and no CRM is connected, both score as unknown and the summary says fit is partly unknown.

### 4. Assign the lead tier
- **A**: fit 30+ and engagement 30+ with at least one buying signal (buying-now poll answer, product question or sales CTA click); or a talk-to-sales or demo CTA click with fit 20+; or a buying-now poll answer from a target account.
- **C**: total under 40; or watch share under 25% with no poll, question or click.
- **B**: everything else.
- Write `score_reasons` (2 to 4 short reasons). Check the A share; if over about a third of scored attendees, show counts and ask.

### 5. Sales notes and talk tracks
For every A and B lead:
- `sales_note`: one line summarizing what they asked or answered, in plain words. "Asked whether the CRM sync handles custom objects" not "Q: does ur sync do custom obj??".
- `talk_track`: two sentences the owner can open with, tied to the question or poll answer, ending with a question. Don't invent product claims; if the right answer isn't known, say "confirm with SE before the call".
- If the question was answered live, say so in the note, so sales doesn't repeat it.

### 6. Route and set next steps
Owner order: open opportunity owner, account owner, then round-robin or territory. A: call or personal email within one business day. B: personalized email within two business days with the resource tied to their question. C: standard post-webinar email or nurture if consent allows.
Set `campaign_member_status` to Attended (live) or Attended On-Demand (replay only).

## Script (optional)

`scripts/score_webinar.py` applies the rubric with the standard library once columns are mapped to the expected names (see its header):

```
python scripts/score_webinar.py attendees.csv --out scored.csv --session-minutes 45 --webinar-date 2027-04-22 \
  --hot-poll "Evaluating now,Next 3 months" --sales-cta "Book a demo" \
  --icp-departments Marketing,Sales --icp-industries Software,Healthcare --target-domains stackline.example.com
```

It doesn't write sales notes or talk tracks; do those in the conversation. Without code execution, apply the rubric directly in batches of 100.

Try it with `assets/sample-attendee-report.csv` (session length 45 minutes).

## Output

**1. Summary**: rows in, attendees scored, no-shows moved out, excluded with reasons, tier counts (A, B, C), live vs on-demand split, top three questions themes, and assumptions.

**2. Ranked handoff list** (CSV, and Markdown for the A list), sorted by tier then total score:
`email`, `first_name`, `last_name`, `company`, `title`, `attendance_type`, `watch_minutes`, `watch_share`, `poll_answers`, `questions`, `cta_clicks`, `engagement_score`, `fit_score`, `total_score`, `lead_tier`, `score_reasons`, `sales_note`, `talk_track`, `owner`, `next_step`, `due_date`, `campaign_member_status`, `flags`.

**3. Question digest**: every question asked, grouped by theme, with who asked it. Useful for the FAQ, the next webinar and product marketing.

An example is in [references/examples.md](references/examples.md).

## Validation checklist

- [ ] Rows in = people scored + excluded + no-shows moved out.
- [ ] Every A and B lead has a sales note, talk track, owner and due date.
- [ ] Sales notes and talk tracks only use what the person asked or answered.
- [ ] Watch share never exceeds 100% (cap it; replays can loop).
- [ ] Campaign member status is Attended or Attended On-Demand for every scored person.
- [ ] EU/UK attendees without opt-in aren't routed into marketing nurture.

## Approval gates

- Before creating CRM tasks, updating lead scores or owners, or setting statuses through a connector, show a preview of every change and get an explicit yes. Write a small test batch first.
- Default to returning files. Never send emails or post to Slack.

## Privacy

Attendee engagement data is personal data. Keep it in the user's tools; don't look people up on the web. Chat messages can contain personal remarks: use only business-relevant content in notes.

## Pushback

- "Weight watch time only": possible, but watch time alone rewards people who left the tab open. Keep at least questions and polls in the score.
- "Push everyone over 50% watch time to sales": show how many that is and how few of them asked or clicked anything; suggest B with a fast email instead.

## Related skills

no-show-recovery (registrants who didn't watch), event-lead-scoring (same tiers for in-person leads), event-follow-up-sequences (emails per lead tier), crm-event-import (Attended and Attended On-Demand statuses).
