Trade show lead research agent for small B2B SaaS sales teams on HubSpot

What The Researcher does

A 11 to 50 person B2B SaaS company exhibits at several trade shows a year, comes home with hundreds of badge scans and business cards, and then watches those leads sit untouched for weeks while the small sales team catches up on existing pipeline. By the time anyone follows up, the prospect has forgotten the booth conversation and may already be in a trial with a competitor. The Researcher owns the gap between the badge scanner and the first rep touch. It works for the sales team (account executives, SDRs and the sales lead). It cleans and dedupes the imported list, enriches each contact and company, checks HubSpot for existing relationships, trial signups and product usage, writes a short account brief and a deep-research brief for top-tier accounts, attaches the relevant competitive battlecard, scores and tiers every lead, and drafts a first touch that references the event. It also maintains the battlecard library as competitor pages and win/loss notes change. It does not send outreach on its own or change deal stages. Reps review and send, and the agent escalates whenever data is thin or conflicting.

Built on HubSpot

The team already runs HubSpot, so the Researcher should live there. HubSpot's Breeze prospecting and company research capabilities handle account summaries and record enrichment natively, and briefs, tasks, scores and sequences sit where reps already work, which matters for a small team with no time to check another tool. Breeze falls short on deep multi-source research, waterfall contact enrichment, product usage joins and structured battlecard upkeep, so those pieces run as a custom AI build orchestrated in n8n, calling Apollo, FullEnrich and Warmly and writing results back to HubSpot through its API. This hybrid keeps the rep experience inside HubSpot while letting the research quality go beyond what a single built-in agent can do, and because the team wants it maintained for them, the n8n workflows and prompts are versioned and monitored so breakage is caught before a show.

What starts it

• A trade show lead list (badge scan CSV or event app export) is uploaded to HubSpot and tagged with the event name
• A new contact is created with lead source set to an event or field marketing campaign
• A rep requests a deep-research brief from a contact or company record before a call or follow-up
• An event lead starts a free trial, revisits the pricing page or is identified as a returning site visitor
• A battlecard is older than its review date, or a competitor is mentioned in a call note or closed-lost reason

What it reads

• HubSpot CRM (contacts, companies, deals, event campaigns, past emails, notes and closed-lost reasons)
• Product usage and trial data synced into HubSpot company properties (signups, active users, feature adoption, plan tier)
• Enrichment from Apollo with FullEnrich as a waterfall for missing emails and direct dials
• Website visit and buying signals from Warmly (or RB2B for person-level identification)
• Public web sources: company site, careers page, press releases, funding news, review sites and competitor pricing and feature pages
• Call transcripts and notes from Fireflies, plus the team's internal win/loss notes for battlecard updates

When a person takes over

• Rep review before any outreach: the agent drafts emails and sequences, but a rep reads, edits and sends every Tier 1 and Tier 2 touch, because booth context and tone matter more than speed alone.
• Existing relationship conflicts: when an event lead matches an open deal, a current customer or an active trial owned by someone else, the agent stops and notifies the record owner instead of creating a new follow-up.
• Thin or conflicting data: if enrichment cannot confirm the company, role or email, or sources disagree, the lead is placed in a review queue for an SDR rather than scored on a guess.
• Battlecard changes: proposed updates to competitive claims, pricing comparisons or positioning go to the sales lead (and product marketing if present) for approval before reps see them.

Rules it follows

• Never send email, enroll sequences, change deal stages or overwrite rep-entered fields without approval; the agent writes to its own properties and tasks.
• Every factual claim in a brief or battlecard cites its source and date, and competitor statements are limited to publicly verifiable information. If the agent cannot verify something, it says so instead of filling the gap.
• SaaS-specific: product usage data is used to tailor the conversation, never quoted back to the prospect in a way that feels like surveillance (no 'we saw you logged in 6 times'). Trial users already in a product-led onboarding flow are coordinated with that flow, not double-touched by sales.
• Respect consent and region rules: only contacts who scanned at the booth or have a lawful basis get sales follow-up, opt-outs and unsubscribes in HubSpot are always honored, and EU contacts follow the stricter path set by the team.
• Event follow-up copy references only what actually happened (the event name, booth notes); it never invents a conversation or claims a demo took place.

How you know it works

• Median time from list upload to first rep touch on Tier 1 leads
• Share of event leads with a completed brief and tier within 48 hours of upload
• Meetings booked from each trade show, by tier
• Pipeline created from event leads within 30, 60 and 90 days of the show
• Rep edit rate on agent drafts and briefs (a falling rate signals quality is improving)

From pilot to live

1. Pilot on one past event (weeks 1 to 3)Run the agent on the most recent show's list that was never fully worked. Confirm HubSpot properties, scoring logic and brief format with two reps, build the first three battlecards for top competitors, and compare the agent's tiers with rep judgment to tune the model.
2. Live on the next trade show (weeks 4 to 8)Trigger the full workflow on the next event upload. Reps send Tier 1 follow-ups from agent drafts within two working days, Tier 2 sequences run with approval, and re-engagement monitoring is switched on. Review KPIs and edit rates weekly and adjust prompts, tiers and enrichment rules.
3. Always-on research for all inbound and events (month 3 onward)Extend the agent to webinar registrants, inbound demo requests and high-intent trial signups, add on-demand deep-research briefs from any record, and move battlecard maintenance to a weekly cadence. Ongoing maintenance covers enrichment credit checks, prompt tuning, HubSpot property changes and quarterly scoring reviews.