Turn an event website into a trade show meeting list

Start in Airtable with a short event prospecting brief before scraping anything. Create fields for event name, event URL, target meeting count, target personas, ICP fit criteria, excluded company types, priority industries, territories, rep owners, and event-specific value proposition. Decide what qualifies as a worthwhile meeting, such as buying committee access, strategic account fit, active session participation, or a known initiative tied to your offer. QA check: have field marketing and sales agree on the qualification rules so the list does not become every logo from the event site.

Create an event prospecting brief for {{event_name}}.

Event URL:
{{event_url}}

Our ICP:
{{icp}}

Target personas:
{{target_personas}}

Territories or regions:
{{territories}}

Offer or conversation topic:
{{offer_or_topic}}

Output:
1. Meeting goal
2. ICP fit criteria
3. Disqualifiers
4. Priority event signals
5. Suggested account tiers
6. Buyer roles to source
7. Meeting invite angles to look for
8. QA rules before outreach

Keep this practical for field marketing and SDRs.

Use Firecrawl to collect the pages that usually contain meeting targets: exhibitors, sponsors, speakers, agenda, session pages, partner pages, and event app previews if they are public. Use crawl mode for event directories and scrape mode for individual pages where you need cleaner markdown or structured extraction. Save the source URL, page type, extraction date, and raw page text for each row so you can verify where every company came from. QA check: spot-check at least 20 extracted rows against the live event website before enrichment.

Clean the Firecrawl export into one Airtable table with one company per row. Add fields for company name, domain, event role, sponsor tier, booth number, speaker name, speaker title, session topic, source page, source URL, and notes from the event page. Use Claude to dedupe name variants, separate vendors from prospects, and flag records that need manual review. QA check: do not enrich rows with missing company names, unclear domains, duplicate domains, or companies classified as competitors or event vendors.

Clean and normalize this event data.

Raw event rows:
{{raw_event_rows}}

Our ICP:
{{icp}}

For each company, return:
1. Clean company name
2. Likely domain if available
3. Event role: sponsor, exhibitor, speaker company, partner, attendee target, competitor, or vendor
4. Sponsor tier or booth info
5. Session topic or speaker context
6. Source URL
7. Data confidence: high, medium, or low
8. Manual review reason if needed

Deduplicate obvious company name variants and keep source evidence.

Send the normalized account table into Clay and enrich each company with domain, employee count, industry, headquarters, LinkedIn company page, funding or growth signals where available, and CRM-matching fields if you use them. Build a scoring model with fit, event relevance, urgency, meeting likelihood, territory ownership, and disqualification reason. Classify each account as high priority, nurture, partner, competitor, skip, or manual review. QA check: manually inspect the top 25 scored accounts and tune the scoring before sourcing contacts.

Score these event accounts against our ICP.

ICP:
{{icp}}

Event context:
{{event_context}}

Enriched account rows:
{{enriched_event_accounts}}

For each account, return:
1. Fit score from 1-5
2. Event relevance score from 1-5
3. Meeting likelihood score from 1-5
4. Priority classification: high priority, nurture, partner, competitor, skip, or manual review
5. Best reason to meet
6. Disqualification reason if applicable
7. Evidence from the event data

Be conservative. Do not mark accounts high priority without a clear fit reason.

Use Apollo, Clay, and Sales Navigator to find 2-4 contacts at each high-priority account. Prioritize people who are listed as speakers, attending the event, owning the relevant function, connected to the session topic, or likely to influence the business problem your offer solves. Capture name, title, LinkedIn URL, email if available, account role, likely buying role, event connection, and confidence level. QA check: remove recruiters, students, agencies, vendors, competitors, generic inboxes, and contacts outside the target geography or function.

Before outreach, match target companies and contacts against HubSpot. Add fields for CRM owner, lifecycle stage, open deal status, customer status, recent outreach, last meeting date, active sequence, and suppression reason. Route existing customers, active opportunities, and named accounts to the correct owner instead of letting a generic event sequence touch them. QA check: no invite should be approved until account ownership and suppression status are filled in.

Use Claude to create one meeting angle per priority contact based on the account fit score, event role, session topic, company signal, buyer role, and your value proposition. The angle should explain why a conversation at this event makes sense now, not simply mention that both companies are attending. Store the angle, supporting evidence, confidence level, and suggested CTA in Airtable. QA check: reject any angle that would still work if you removed the event name, because that means it is not event-specific enough.

Create event-specific meeting angles for these contacts.

Event context:
{{event_context}}

Account and contact rows:
{{priority_contacts}}

Our value proposition:
{{value_proposition}}

For each contact, return:
1. Meeting angle under 50 words
2. Why this angle fits
3. Event evidence used
4. Buyer role relevance
5. Suggested CTA
6. Confidence level
7. What the rep should verify before sending

Avoid generic 'are you attending?' language.

Turn the approved meeting angles into one short email and one LinkedIn note per contact. Include the event name, the specific relevance hook, one conversation topic, a soft CTA, and optional meeting windows if reps already have availability. Keep each message short enough that a busy attendee can understand the ask in under 15 seconds. QA check: review the first 20 messages for unsupported claims, fake personalization, wrong persona assumptions, and overly aggressive demo language.

Draft an event meeting invite email and LinkedIn note for each contact.

Contacts and angles:
{{contact_angle_table}}

Event name:
{{event_name}}

Meeting windows:
{{meeting_windows}}

Rules:
- Email under 120 words
- LinkedIn note under 500 characters
- Mention one specific event or session context
- Include one conversation topic
- Use a soft CTA
- Do not overclaim, flatter, or pretend we know private information

Return email subject, email body, LinkedIn note, and QA warning if any.

Before sending the full list, audit the first 20 invites across different account tiers and reps. Check account fit, contact fit, event evidence, CRM ownership, message relevance, CTA strength, and whether the outreach sounds like a real person wrote it. Use Claude to flag patterns, but let the campaign owner make the final approval decision. QA check: fix the scoring, segmentation, or prompt before scaling if more than 20 percent of the sample needs edits.

Audit these event meeting invites before launch.

Invite sample:
{{invite_sample}}

ICP and event brief:
{{event_prospecting_brief}}

Check for:
1. Bad fit
2. Unsupported personalization
3. Weak event relevance
4. Wrong persona assumption
5. Vague CTA
6. Duplicate CRM ownership risk
7. Copy that sounds automated

Return pass, revise, or reject for each invite, plus the specific fix.

Assign every approved contact to a rep in Airtable or HubSpot and create outreach waves for 3-4 weeks before the event, one week before, during the event, and post-event. Add owner, send date, follow-up date, meeting window, channel, status, and notes for each contact. Keep the waves small enough that reps can personalize, respond, and book meetings without losing track. QA check: every high-priority account needs one owner, one next action, and one deadline before launch.

Send the approved invites through the rep's normal email or LinkedIn workflow rather than a blind bulk blast. Track sent, opened if available, replied, booked, declined, no response, follow-up needed, and meeting completed in Airtable or HubSpot. Review booked meetings by account tier, persona, event signal, and rep owner so you know what is creating real pipeline. QA check: pause the campaign if negative replies, duplicate outreach, or wrong-account complaints appear.

After the event, segment contacts into met live, booked but no-show, replied but not met, no response, new contact added at booth, and disqualified. Draft follow-up by segment while the event context is still fresh, and update HubSpot with meeting notes, next steps, and campaign attribution. Use the performance data to identify which event signals produced meetings: sponsor tier, session topic, speaker status, account tier, or persona. QA check: every booked meeting should have a next step or a closed-loop disqualification within one week of the event.

Segment these event prospects and draft post-event follow-up messages.

Prospect status table:
{{prospect_status_table}}

Event recap:
{{event_recap}}

Offer or next step:
{{offer_or_next_step}}

Create segments for:
- Met live
- Booked but no-show
- Replied but not met
- No response
- New contact added at booth
- Disqualified

For each segment, write a short follow-up message, CRM update note, and recommended next action.