The ICP Qualifier
A prompt that tells you in 10 seconds if a company is worth emailing. Upload your list, get back MATCH, NO MATCH, or INSUFFICIENT DATA for every row.
What This Does
You have a spreadsheet of companies exported from Sales Navigator, Apollo, LinkedIn, or any data provider. You want to know which ones are actually worth emailing – without opening 2,000 tabs.
This prompt turns Claude into a qualification engine. Upload your CSV, and it evaluates every company against your ICP criteria using the data already in your spreadsheet. You get back the same CSV with three new columns: Verdict (MATCH, NO MATCH, or INSUFFICIENT DATA), Reason (a short explanation), and Failed Check (which step caught it).
How It Works (3 Steps)
1 Define Your ICP
Copy the template below and fill in every field. The quality of your output depends entirely on the quality of your ICP definition. Be specific, not aspirational.
WHO I SELL TO: [Describe your ideal customer in one paragraph. Include their industry, what they do, and the problem you solve for them. Example: "B2B SaaS companies with outbound sales teams who need help booking more meetings through cold email and LinkedIn outreach."] WHAT MY IDEAL CUSTOMER MUST HAVE: [List 2-4 things a company needs to be a fit. Example: "Must have a dedicated sales team or SDR function. Must sell to other businesses. Must be actively doing outbound or planning to start."] MINIMUM COMPANY SIZE: [Number of employees, or "no minimum." Example: "25+ employees"] EXCLUDED INDUSTRIES: [Industries that are never a fit, even if they match everything else. Example: "Recruiting/staffing agencies, non-profits, government, education."] FALSE POSITIVES - COMPANIES THAT LOOK LIKE A FIT BUT AREN'T: [This is the most important field. List 3-5 specific types of companies that would pass a surface-level check but are NOT your customer. These are the ones that waste your time. Example: "Marketing agencies that do social media for restaurants - they look like B2B but their clients are consumer businesses. CRM consultants who only do implementation, not sales strategy. HR tech companies that have 'sales' in their description but sell to HR departments, not revenue teams."] B2B2C POLICY: [Do you want to exclude companies that sell to businesses whose end customers are consumers? This catches companies like: tools for car dealerships (dealers sell to retail buyers), lead gen for insurance brokers (brokers sell to individuals), marketing platforms for real estate agents (agents sell to homebuyers), appointment software for dental practices (patients are consumers). Answer: "Exclude B2B2C" or "Include B2B2C"]
2 Prep Your CSV
Export your prospect list as a CSV. At minimum, you need a column with the company name and a column with a company description (LinkedIn company description works). If you have industry and employee count columns, include them.
3 Run the Qualifier
Open a new Claude conversation. Paste the prompt below, replace the [MY ICP DEFINITION] placeholder with your filled-in ICP from Step 1, upload your CSV, and hit send.
You are an ICP Qualification Engine. Your job is to evaluate every company in the uploaded CSV and determine whether it is a MATCH, NO MATCH, or INSUFFICIENT DATA against my Ideal Customer Profile. ## MY ICP [PASTE YOUR ICP DEFINITION FROM STEP 1 HERE] ## EVALUATION RULES Work through these checks in order for each company. Stop at the first failure. ### Check 1: Data Quality Evaluate the available company information (description, industry, any other fields). If the description is empty, contains only a company name, or is a generic tagline under 10 words with no specifics about what the company does (e.g. "Innovative solutions for modern businesses"), mark as INSUFFICIENT DATA with reason "Not enough information to evaluate. Review manually." Do not guess. If a homepage text column is present, use it as the primary evaluation source. It is far more reliable than LinkedIn descriptions. If both homepage text and a description are present, prioritize homepage text. ### Check 2: B2B Verification The company must sell products or services to other businesses. If the company primarily sells to individual consumers (retail, DTC, personal finance, personal health, entertainment, food and beverage to consumers), mark as NO MATCH. Failed check: "B2B Verification." ### Check 3: B2B2C Filter If my ICP definition says "Exclude B2B2C," apply this check. A company is B2B2C if it sells to businesses whose end customers are individual consumers. The test: look at who the company's customers ultimately sell to. If the end buyer is a person, not a business, the company is B2B2C. Common B2B2C examples that must be caught: - Tools for car dealerships (dealers sell to individual buyers) - Lead gen or CRM for insurance brokerages (brokers sell to individuals) - Marketing platforms for real estate agents (agents sell to homebuyers) - Software for financial advisors (advisors serve individual clients) - Appointment or scheduling tools for dental practices, med spas, salons, gyms (patients/clients are consumers) - POS, loyalty, or marketing tools for restaurants, retail stores, or hospitality (end customers are consumers) - Home service provider tools - HVAC, plumbing, roofing, landscaping (homeowners are consumers) If the company serves these types of businesses, mark as NO MATCH with reason describing the B2B2C relationship. Failed check: "B2B2C Filter." ### Check 4: False Positive Filter Check the company against the false positive examples in my ICP definition. These are company types that look like a fit on the surface but are explicitly excluded. If the company matches any of these patterns, mark as NO MATCH. Failed check: "False Positive." ### Check 5: Industry Exclusion If the company operates in an excluded industry from my ICP definition, mark as NO MATCH. Failed check: "Industry Exclusion." ### Check 6: Company Size If employee count data is available AND my ICP specifies a minimum, check it. If the company falls below the minimum, mark as NO MATCH. If employee count data is not available, skip this check (do not fail a company for missing data). Failed check: "Company Size." ### Check 7: ICP Alignment Evaluate whether the company's primary product or service aligns with what I defined in "Who I sell to" and "What my ideal customer must have." IMPORTANT: Do not give the benefit of the doubt. A company must show clear evidence of alignment based on the available data. Vague or ambiguous descriptions that could fit many interpretations should not be treated as a match. If you cannot confidently determine alignment, mark as INSUFFICIENT DATA. Evaluate what the company actually does, not just the words they use. Marketing language often obscures the real service. A company that says "we help businesses grow" could be anything. Look for specifics: what is the product, who uses it, and what outcome does it drive. If the company passes all checks, mark as MATCH. Failed check: N/A for matches. ## AMBIGUITY RULES - When in doubt, mark INSUFFICIENT DATA rather than guessing MATCH or NO MATCH. - A company that "might" fit is not a MATCH. You need clear evidence. - If a description mentions multiple services, evaluate based on the primary service, not a secondary offering. - If a company could serve both B2B and B2C markets and you cannot determine the primary focus, mark INSUFFICIENT DATA. ## OUTPUT INSTRUCTIONS Return a CSV file that contains ALL original columns from the uploaded CSV, plus three new columns appended at the end: 1. **Verdict** - exactly "MATCH", "NO MATCH", or "INSUFFICIENT DATA" 2. **Reason** - one sentence explaining why (max 15 words) 3. **Failed Check** - which check caught it (e.g. "B2B2C Filter", "ICP Alignment", "Data Quality") or "N/A" for matches Preserve every row. Do not skip, reorder, or modify any original data. Output the complete CSV as a downloadable file. Process every row. If the CSV has 500 rows, the output CSV must have 500 rows.
What to Do With the Output
You now have three groups:
MATCH – Your qualified list. Enrich with contact data, load into your outreach tool, or push to Clay for further processing. These are worth your time.
NO MATCH – Check the Failed Check column. If most are failing on the same check, your filters are working. If companies are failing that shouldn’t be, tighten your ICP definition and re-run.
INSUFFICIENT DATA – These need a manual look. Open the company’s website or LinkedIn page, get more information, and make a judgment call. This is a much smaller list than your original pull, which is the point. The prompt did the easy decisions so you only spend time on the hard ones.
Example
Say you sell a sales engagement platform to B2B SaaS companies. Your ICP excludes B2B2C and recruiting/staffing firms. Here’s how the qualifier handles a few rows:
Company: Outreach.io Description: "Sales execution platform that helps B2B sales teams create and close more pipeline." Verdict: MATCH Reason: B2B sales execution platform, strong ICP fit. Failed Check: N/A Company: DealerSocket Description: "CRM and digital marketing solutions for automotive dealerships." Verdict: NO MATCH Reason: Serves car dealerships selling to consumers. Failed Check: B2B2C Filter Company: Hirewell Description: "Recruiting and talent solutions for growing companies." Verdict: NO MATCH Reason: Recruiting/staffing, excluded industry. Failed Check: Industry Exclusion Company: TechVentures LLC Description: "Innovative solutions for modern businesses." Verdict: INSUFFICIENT DATA Reason: Generic description, cannot verify alignment. Failed Check: Data Quality Company: LeadSpark Description: "AI-powered lead generation for insurance agents and financial advisors." Verdict: NO MATCH Reason: B2B2C - agents/advisors serve consumers. Failed Check: B2B2C Filter
B2B Leads That Convert: The weekly newsletter for B2B sales professionals who prospect.