---
name: event-lead-scoring
description: Scores and routes leads from trade shows, conferences, field events and webinars into lead tier A, B or C using booth notes, qualifying-question answers, ICP fit, existing pipeline and account ownership, then assigns an owner, next step and due date per lead. Works from a CSV or pasted list, or reads accounts and owners from HubSpot or Salesforce when connected. Use for event lead scoring, lead routing, sales handoff, a hot lead list for Slack or deciding who follows up with whom after an event.
license: MIT
metadata:
  author: eventmarketingsoftware.com
  version: "1.0"
  category: leads-and-follow-up
  source: https://eventmarketingsoftware.com/skills/event-lead-scoring
---

# Event lead scoring and routing

Sort event leads into lead tier A, B or C with a transparent score, then give every lead an owner, a next step and a due date. The goal is a short list sales will actually work and a definition of a qualified event lead marketing can defend.

Terms, lead tier definitions, flags and timing follow [references/shared-terms.md](references/shared-terms.md). Input is ideally the clean file from badge-scan-cleanup (and notes from booth-notes-to-crm). Output feeds event-follow-up-sequences and crm-event-import.

## Intake

1. Read `event-context.md` (or `event-program-context.md`) if the user has one: ICP, tier rules, territories, owners, target account list, CRM.
2. Ask for anything still missing, in one message:
   - The lead list (CSV, XLSX, pasted, or a Google Sheet via connector). If it hasn't been cleaned, offer to run the badge-scan-cleanup steps first, since duplicates distort scoring.
   - ICP: industries, company size range, target departments and seniority.
   - The booth qualifying questions and what a "hot" answer looks like.
   - Account ownership: a CRM connector, or an export with domain, account owner, customer yes/no and open opportunity owner.
   - Routing rules: territories, round-robin list or a default owner.
   - Event end date (for due dates).
3. If the user can't give an ICP, use the defaults in [references/scoring-rubric.md](references/scoring-rubric.md) and say so at the top of the output.

## Connectors (optional)

- **HubSpot or Salesforce connector available**: look up each lead's email and domain to find existing contacts or leads, account owner, customer status and open opportunities. Read only. Don't create tasks or change owners until the approval gate below.
- **Slack connector available**: you can post the hot list after approval.
- **Google Sheets connector available**: read the list and write results to a new tab.
- **Otherwise**: work from the pasted or uploaded files and return a CSV plus Markdown summary tables.

## Method

### 1. Prepare
- Confirm the standard columns exist (see shared terms). Map any variants. Keep every input column.
- Set aside leads with `lead_tier = Exclude`: flags competitor, internal, student or test_record. They stay in the file with the reason; they aren't scored.
- Leads flagged `possible_duplicate` (same person, two emails) get scored separately but routed to the same owner; list the pairs in the summary so the owner can merge them.
- Leads flagged `exhibitor_or_vendor` are scored but capped at C unless the notes show they're also a buyer.

### 2. Match to the CRM
For each lead set `crm_match`: `new`, `existing_lead`, `existing_contact`, `existing_account` (the domain matches an account but not the person) or `customer`, plus `account_owner` and `open_opp` (yes/no and owner). Match on email first, then on company domain. Personal-email leads match only on email. Never match on company name alone without saying so.

### 3. Score fit and intent
Use [references/scoring-rubric.md](references/scoring-rubric.md):
- **Fit score (0 to 50)**: seniority, department, company size, industry, target account.
- **Intent score (0 to 50)**: what happened in the conversation (meeting booked or held, demo or pricing request, named project or pain, general interest, scan only), the timeline answer, budget signals and repeat engagement (scan_count above 1, several people from one account).
- Read the notes like a sales manager would. "Wants demo" and "asked about pricing" are intent; "grabbed swag" and blank notes aren't. Note negative signals too ("happy with current tool, contract runs to 2029").

### 4. Assign the lead tier
- **A**: fit 30 or more and intent 30 or more; or a meeting booked or held; or a demo or pricing request with fit 20 or more; or a target account with a real conversation.
- **C**: total under 40, or intent 0 (scan only, no real conversation), however good the fit.
- **B**: everything else.
- Write `tier_reasons` as 2 to 4 short plain reasons a rep would accept, e.g. `VP Marketing; 0-3 month timeline; asked for demo`.
- If the user's tier rules differ, use theirs. If you override the rubric for one lead, fill `override_reason`.
- Sanity check the distribution. If more than about a third of scored leads are A, the bar is probably too low; show the counts and ask before continuing.

### 5. Route
Pick an owner in this order and record `owner_source`:
1. Open opportunity owner (`open_opp`), and mark the lead as influenced, not sourced.
2. Existing account owner (`account_owner`). Customers go to the account manager or CSM unless the notes show a new buying project.
3. For A leads with no owner, the rep who had the conversation, if they carry a quota.
4. Territory rules (country, state, segment).
5. The owner already given to another lead from the same company domain.
6. Round-robin across the list the user gave (balanced by A leads first), or the default owner.
Details and edge cases are in [references/routing-rules.md](references/routing-rules.md).

### 6. Next step and due date
- A: `next_step` is specific to the notes ("Call to confirm Thursday demo and invite ops lead"). Due within one business day of the event end (or of the conversation for multi-day shows).
- B: personalized email with the asset they asked about. Due within two business days.
- C: one light-touch email, then nurture if consent allows. Due within five business days.
- Leads flagged `eu_uk_no_optin`: next step can be a one-to-one reply about the conversation, never a marketing list add. Say "check consent before nurture".

## Script (optional)

`scripts/score_leads.py` applies the rubric deterministically with the standard library:

```
python scripts/score_leads.py leads.csv --out scored.csv --event-end 2027-03-11 \
  --accounts accounts.csv --target-domains stackline.example.com,meadowbank.example.com \
  --icp-departments Marketing,Sales --icp-industries Software,Logistics,Healthcare \
  --icp-min-employees 200 --icp-max-employees 5000 --owners "Alex Kim,Jo Park"
```

It scores intent from note keywords, so always read the A and borderline B leads (total 35 to 45) yourself and adjust with an `override_reason`. Without code execution, apply the same rubric in the conversation in batches of 100 rows.

Try it with `assets/sample-leads.csv` and `assets/sample-accounts.csv`.

## Output

**1. Summary tables**

| Lead tier | Leads | Share | Owners |
|---|---|---|---|
| A | | | |
| B | | | |
| C | | | |
| Exclude | | | |

| Owner | A | B | C | Due first |
|---|---|---|---|---|

Plus: rows in and rows out, how many matched the CRM, how many are influenced (open opportunity) vs new, and any assumptions.

**2. Scored lead file**: all input columns, then `lead_tier`, `total_score`, `fit_score`, `intent_score`, `tier_reasons`, `crm_match`, `account_owner`, `open_opp`, `owner`, `owner_source`, `next_step`, `due_date`, `override_reason`. Sort by tier, then total score descending.

**3. Hot list (optional, for Slack or email)**: top A leads, one line each:

```
*Example Expo 2027: A leads (12)*  Follow up by Fri Mar 12
1. Dana Reyes, VP Marketing, Brightpath. Asked for a lead routing demo, budget approved for Q3. Owner: Alex Kim. Next: confirm demo, invite ops lead Priya Shah.
2. ...
Full list: <where the file lives>
```

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

## Validation checklist

- [ ] Every input row appears in the output exactly once with a lead tier (A, B, C or Exclude).
- [ ] Every A and B lead has an owner, a next step and a due date.
- [ ] Every Exclude has a reason.
- [ ] Leads with an open opportunity went to the opportunity owner.
- [ ] No EU/UK lead without opt-in has a next step that adds them to marketing email.
- [ ] Tier counts add up to the total and the A share looks plausible.

## Approval gates

- Before creating CRM tasks, changing lead or contact owners, or setting fields in HubSpot or Salesforce, show a preview table of exactly what will change (record, field, old value, new value) and get an explicit yes. Write in small batches and report results.
- Before posting the hot list to Slack, show the exact message and channel and get a yes. Don't post personal phone numbers or emails to a channel; names, companies and next steps are enough.
- Default to returning files and drafts.

## Privacy

Lead data is personal data. Keep it in the user's own tools. Don't look people up on the web or enrich from outside sources. Keep flags like `eu_uk_no_optin` visible to the owner.

## Pushback

- "Make them all A": explain that A means a rep acts within a day, and an inflated A list gets ignored. Offer a bigger B tier with a fast sequence instead.
- "Score without notes": possible, but intent will rest on qualifier answers and scan_count only. Say so in the summary.

## Related skills

badge-scan-cleanup (before), booth-notes-to-crm (adds notes), event-follow-up-sequences (writes the emails per lead tier), crm-event-import (campaign member status and import), webinar-attendee-scoring (the same idea for webinar engagement data).
