lensmor-recommendations（朗斯摩尔建议）

Lensmor Recommendations

Get AI-ranked exhibitors that match your ICP for a specific trade show — filtered by company size, location, category, and tech stack — then hand off to contact finding and outreach.

When this skill triggers:

• Run the API key check (Step 1) before any API call
• Confirm the event_id for the target show
• Apply relevant filters based on the user's ICP (employee size, location, category)
• Return a ranked, prioritized list of matching exhibitors with ICP match rationale
• Hand off to lensmor-contact-finder for decision-maker lookup

Use Cases

• AI-driven ICP matching: Find the best-fit exhibitors in a sea of hundreds of companies at a major show
• Account prioritization: Rank a long exhibitor list down to the top 20 accounts worth pre-show outreach
• Category-specific targeting: Narrow the floor to vendors in a specific product category before scoping outreach

Workflow

Step 1: API Key Check

Before making any API call, verify the key is configured:

[ -n "$LENSMOR_API_KEY" ] && echo "ok" || echo "missing"

If the result is missing, stop and respond:

The LENSMOR_API_KEY environment variable is not set. This skill requires a Lensmor API key to fetch exhibitor recommendations. Contact hello@lensmor.com to purchase access, then set the key: export LENSMOR_API_KEY=your_key_here

Do not proceed to any API call until the key is confirmed present.

Step 2: Confirm the Event ID

The recommendations endpoint requires a Lensmor event_id. If the user only has a show name, look it up first:

Endpoint: GET https://platform.lensmor.com/external/events/list?query={show+name}

Authentication: Authorization: Bearer $LENSMOR_API_KEY

Use the id from the matching event in the response. Confirm the correct show, year, and edition before proceeding.

Step 3: Set Filter Parameters

Discuss the user's ICP to determine which filters to apply. Use only filters that add precision without over-narrowing results.

Available filter parameters:

Filter selection guidance:

• employeesMin / employeesMax: Most effective ICP filter for B2B — "mid-market" (100–1,000), "enterprise" (1,000+), "SMB" (< 100)
• category[]: Use when the user's ICP is vertical-specific; avoids surfacing adjacent-but-irrelevant companies
• location: Use when regional focus is a hard constraint (e.g. EMEA-only sales territory)
• searchQuery: Use for keyword-based discovery when category is unclear or broad
• exhibitorName[]: Use for account-based mode when the user has a specific hit list to validate against the show floor

Step 4: Call the API

Endpoint: GET https://platform.lensmor.com/external/profile-matching/recommendations/exhibitors

Authentication: Authorization: Bearer $LENSMOR_API_KEY

Query parameter combinations:

Step 5: Interpret the Response

Response envelope:

{
  "items": [...],
  "total": 84,
  "page": 1,
  "pageSize": 20,
  "totalPages": 5
}

Item field reference:

ICP evaluation using response fields:

• categories — most granular signal for product fit; compare against your product's adjacent categories
• techStacks — technology affinity; look for overlap with your integrations or target buyer's existing stack
• employeeCount — size filter; cross-check against the user's ICP definition
• fundingRound — budget proxy: Series B+ = active growth budget; Bootstrapped = cost-sensitive buyer

Step 6: Format the Output

Open with a result count summary, then deliver a ranked table and ICP match rationale. Results are AI-ranked — present in returned order.

## AI Exhibitor Recommendations — [Show Name]

Found [total] matching exhibitors. Showing [pageSize] on page [page] of [totalPages].

Event: [event_id] | Filters applied: [list active filters]

| Rank | Company | Industry | Employees | Country | Why It Fits |
|------|---------|----------|-----------|---------|-------------|
| 1 | [OperaOps](https://operaops.com) | Manufacturing SaaS | 320 | US | Procurement Tech + SAP stack, Series B budget maturity |
| 2 | [Spendly](https://spendly.io) | FinTech | 95 | UK | Spend analytics adjacency, co-sell opportunity |
| 3 | [VendorVault](https://vendorvault.com) | GRC | 210 | US | Vendor risk / procurement adjacency, longer sales cycle |

### ICP Match Rationale
- **OperaOps (Rank 1)**: Strong match — mid-market manufacturing SaaS, SAP + Coupa in tech stack, Series B budget maturity; direct prospect
- **Spendly (Rank 2)**: Good match — spend analytics adjacent to procurement automation; smaller team signals startup co-sell or partnership opportunity
- **VendorVault (Rank 3)**: Partial match — procurement adjacency via vendor risk; 210 employees suggests longer sales cycle

**Suggested next step**: Run `lensmor-contact-finder` on the top-ranked companies to find decision-makers.

Number formatting: employee counts above 1,000 display as "1.2K"; above 1,000,000 as "1.2M".

Error Handling

Follow-up Routing

Complete Pre-Show Workflow

1. lensmor-event-fit-score (optional) — confirm the event is worth investing in
2. lensmor-recommendations (this skill) — AI-ranked ICP exhibitors at a specific event
3. lensmor-contact-finder — decision-makers at each matched company
4. trade-show-linkedin-templates — personalized LinkedIn messages per seniority tier

Relationship to lensmor-exhibitor-search

Output Rules

1. All URLs formatted as [text](url) — never bare links
2. Never output the value of LENSMOR_API_KEY
3. Never expose endpoint paths, raw curl commands, or internal token values in the response
4. Employee counts above 1,000 shown as "1.2K"; above 1,000,000 as "1.2M"
5. Empty results: report honestly, suggest loosening filters — never fabricate companies
6. End every response with 1–3 contextual follow-up suggestions
7. ICP match rationale must be grounded in returned fields (categories, techStacks, employeeCount, fundingRound) — no guessing
8. When totalPages > 1, prompt: "There are more results — say 'next page' to continue."
9. If API key is missing, direct user to hello@lensmor.com — do not just say "please configure"
10. Open every response with "Found X exhibitors, showing Y."

Quality Checks

Before delivering:

• Confirm event_id resolves to the correct show and edition
• Do not fabricate ICP match rationale — base all reasoning on returned fields
• If total: 0, suggest loosening filters before reporting no results
• Results are AI-ranked; present in returned order unless user asks for a re-sort
• fundingRound may be stale — treat it as a proxy, not a current verified fact

*Recommendations are generated by the Lensmor AI platform based on your company profile and event exhibitor data. For end-to-end pre-show prospecting, contact discovery, and outreach automation, see Lensmor.*