Using AI
Templates
Find some ready-to-use templates for your AI prompts
0. Prerequisites
Your table must contain the columns mentioned in the "Input" section of your prompt.
Note : You can save these prompts by clicking on "Save as template" (to reuse it the next time, click on "Import a template").
1. Classify as B2B or B2C
Inputs :
Company Name : /Company name
Company domain : /Company domain
You are tasked with determining the business model of a company based on its online presence and domain.
Identify if the company associated with the domain operates as a B2B, B2C, or both.
1. Begin by visiting the company's official website using the provided domain
2. Review the content on their homepage, services, products, and any available "About Us" or "Company Profile" pages
3. Look for key indicators such as language targeting businesses (B2B) or consumers (B2C). Use terms like "business solutions," "enterprise," "clients," for B2B, and "shop now," "customer service," "retail," for B2C
4. Check any available online press releases, product listings, or service descriptions.
5. Search for the target audience mentioned in their marketing materials. Terms like "partners," "resellers," signify B2B, while "shoppers," "individuals," signify B2C.
6. Reference business directories or industry-specific sites if necessary to confirm business model.
7. If conflicting information is found, cross-reference with reliable business news or reports.
Examples:
- For a page focused on enterprise clients and bulk ordering, categorize as "B2B.
- For a site featuring retail products with shopping cart options, categorize as "B2C."
- If both audience types are evident, categorize as "B2B and B2C."
- If the information is ambiguous or unclear, result as "Cannot determine."
Output : Conclude with only "B2B," or "B2C," or "B2B and B2C," or "Cannot determine" based on findings.
Company Name : /Company name
Company domain : /Company domain
You are tasked with determining the business model of a company based on its online presence and domain.
Identify if the company associated with the domain operates as a B2B, B2C, or both.
1. Begin by visiting the company's official website using the provided domain
2. Review the content on their homepage, services, products, and any available "About Us" or "Company Profile" pages
3. Look for key indicators such as language targeting businesses (B2B) or consumers (B2C). Use terms like "business solutions," "enterprise," "clients," for B2B, and "shop now," "customer service," "retail," for B2C
4. Check any available online press releases, product listings, or service descriptions.
5. Search for the target audience mentioned in their marketing materials. Terms like "partners," "resellers," signify B2B, while "shoppers," "individuals," signify B2C.
6. Reference business directories or industry-specific sites if necessary to confirm business model.
7. If conflicting information is found, cross-reference with reliable business news or reports.
Examples:
- For a page focused on enterprise clients and bulk ordering, categorize as "B2B.
- For a site featuring retail products with shopping cart options, categorize as "B2C."
- If both audience types are evident, categorize as "B2B and B2C."
- If the information is ambiguous or unclear, result as "Cannot determine."
Output : Conclude with only "B2B," or "B2C," or "B2B and B2C," or "Cannot determine" based on findings.
2. Venture-backed?
Inputs :
Company Name : /Company name
Company domain : /Company domain
Search the web to determine if this company has received venture capital or institutional funding.
Follow these rules strictly:
Only confirm venture-backed status if you find a concrete source : a press release, Crunchbase page, TechCrunch article, or official announcement mentioning a funding round with an amount or an investor name.
If you find no concrete source, output "Not confirmed" and nothing else. Do not assume or infer venture-backed status from the company's size, language, or industry.If confirmed, list for each round : round name (Seed, Series A, etc.), amount raised, investors names, date, and source linkIf only partial information is found, list what is confirmed and explicitly flag what is missing with "Unknown".
Output format:Venture-backed : Yes / No / Not confirmed
If Yes : Round / Amount / Investors / Date / Source
Company Name : /Company name
Company domain : /Company domain
Search the web to determine if this company has received venture capital or institutional funding.
Follow these rules strictly:
Only confirm venture-backed status if you find a concrete source : a press release, Crunchbase page, TechCrunch article, or official announcement mentioning a funding round with an amount or an investor name.
If you find no concrete source, output "Not confirmed" and nothing else. Do not assume or infer venture-backed status from the company's size, language, or industry.If confirmed, list for each round : round name (Seed, Series A, etc.), amount raised, investors names, date, and source linkIf only partial information is found, list what is confirmed and explicitly flag what is missing with "Unknown".
Output format:Venture-backed : Yes / No / Not confirmed
If Yes : Round / Amount / Investors / Date / Source
3. Recent company news
Inputs :
Company name : /Company name
Company domain : /Company domain
Scrape the web to find recent news articles for this company. Strictly discard any article or information dated before September 2025. If you cannot confirm the date of an article, discard it.
Search for news in the following categories, listed by priority order:
Fundraising and Investment Rounds (highest priority) : venture capital funding, Series A/B/C rounds, private equity investments, or any mention of securing new capital.
Leadership Changes in Marketing : new CMO, marketing director, brand manager, or VP Marketing appointments.
Expansion into New Markets : new geographical markets, international growth, new store openings, new customer segments.
Marketing Campaigns and Rebranding : new advertising campaigns, rebranding initiatives, product launches, brand repositioning.
Industry Awards and Recognition : marketing awards, brand excellence awards, advertising accolades.
Partnerships : new partnerships, acquisitions, or group ownership connections.
Search instructions: Keep searches simple and do not add quotes. Try multiple formats in this order:
(company name recent news)
(company name news 2025)
(company name news 2026)
(company name funding OR hiring OR expansion OR partnership OR rebrand OR campaign)
If you cannot find any news dated after September 2025, output "Not found" and nothing else.
Output format:
Provide results structured as follows:
High priority signals (Funding, Leadership)
[bullet points]
Medium priority signals (Expansion, Campaigns)
[bullet points]
Low priority signals (Awards, Partnerships)
[bullet points]
Sources:
[numbered links]
Each bullet point must include the date of the article or event in parentheses at the end, formatted as (Month Year) e.g. (March 2026). If the date cannot be confirmed, discard the bullet point entirely.
Company name : /Company name
Company domain : /Company domain
Scrape the web to find recent news articles for this company. Strictly discard any article or information dated before September 2025. If you cannot confirm the date of an article, discard it.
Search for news in the following categories, listed by priority order:
Fundraising and Investment Rounds (highest priority) : venture capital funding, Series A/B/C rounds, private equity investments, or any mention of securing new capital.
Leadership Changes in Marketing : new CMO, marketing director, brand manager, or VP Marketing appointments.
Expansion into New Markets : new geographical markets, international growth, new store openings, new customer segments.
Marketing Campaigns and Rebranding : new advertising campaigns, rebranding initiatives, product launches, brand repositioning.
Industry Awards and Recognition : marketing awards, brand excellence awards, advertising accolades.
Partnerships : new partnerships, acquisitions, or group ownership connections.
Search instructions: Keep searches simple and do not add quotes. Try multiple formats in this order:
(company name recent news)
(company name news 2025)
(company name news 2026)
(company name funding OR hiring OR expansion OR partnership OR rebrand OR campaign)
If you cannot find any news dated after September 2025, output "Not found" and nothing else.
Output format:
Provide results structured as follows:
High priority signals (Funding, Leadership)
[bullet points]
Medium priority signals (Expansion, Campaigns)
[bullet points]
Low priority signals (Awards, Partnerships)
[bullet points]
Sources:
[numbered links]
Each bullet point must include the date of the article or event in parentheses at the end, formatted as (Month Year) e.g. (March 2026). If the date cannot be confirmed, discard the bullet point entirely.
4. List company’s products & services
Inputs :
Company name : /Company name
Company domain : /Company domain
Go to this company's website and identify the main products and services they offer.
Follow these rules strictly:
Only list core offerings : what the company primarily sells or delivers to clients. Ignore internal departments, subsidiaries, buzzwords, or generic marketing language.
Maximum 7 items in the output list
If the company has subsidiaries or business units, list the parent company's core activity, not the subsidiaries as separate services
Deduplicate : if two items mean the same thing, keep only the most specific one
If a service has clear sub-services, group them under one parent label (ex: "Digital Marketing" instead of listing social media, content creation, SEO separately)
Output: a comma-separated list of core products and services, maximum 7 items, no descriptions, no buzzwords.
Capitalize the first letter of each item.
Company name : /Company name
Company domain : /Company domain
Go to this company's website and identify the main products and services they offer.
Follow these rules strictly:
Only list core offerings : what the company primarily sells or delivers to clients. Ignore internal departments, subsidiaries, buzzwords, or generic marketing language.
Maximum 7 items in the output list
If the company has subsidiaries or business units, list the parent company's core activity, not the subsidiaries as separate services
Deduplicate : if two items mean the same thing, keep only the most specific one
If a service has clear sub-services, group them under one parent label (ex: "Digital Marketing" instead of listing social media, content creation, SEO separately)
Output: a comma-separated list of core products and services, maximum 7 items, no descriptions, no buzzwords.
Capitalize the first letter of each item.
5. Determine someone's likely manager
Inputs:
Company name: /Company name
Company domain: /Company domain
Job title: /Job title
First name: /First name
Last name: /Last name
Identify the most likely direct manager of the person above based on the role and company.
Instructions:
Search for the individual's role and company to determine their department and where they fit in the company's hierarchy.
Find others in the same company with similar or related titles to establish the department's reporting structure.
Identify the most likely boss, prioritizing:
Someone one level above in the same department (e.g., a Head of Growth for a Growth Manager).
If no direct manager is found, look for a department leader (e.g., VP of Growth or VP of Marketing for Growth Lead).
If neither is found, look for an executive overseeing the function (but avoid defaulting to the CEO unless they directly oversee the role).
Verify the person works at the company by checking their listed employer against the company domain. Ignore results where the company does not match.
Ensure the LinkedIn profile URL is correctly formatted in the structure "linkedin.com/in/username". Ignore results with different formats (e.g., company pages or incorrect URLs).
Verify the manager exists : confirm the identified manager's name and title appear on their actual LinkedIn profile page before outputting. If you cannot confirm, output "Not confirmed" for that field.
Never output placeholder or generic names : if no real manager is found, output "Not found" and nothing else. Do not invent or approximate a name.
Retrieve their full name, job title, and LinkedIn profile URL.
Output Format:
Boss's Name: [Full Name]
Title: [Job Title]
LinkedIn URL: [linkedin.com/in/username]
Confidence: [High - Medium - Low]
High = LinkedIn profile verified directly, Medium = found via third-party source, Low = inferred without direct verification.
Example Input:
Job title: Growth Manager
First name: Michel
Last name: Bearh
Company name: Acme Inc.
Company domain: acme.com
Example Output:
Boss's Name: Jane Doe
Title: Head of Growth
LinkedIn URL: https://www.linkedin.com/in/janedoe
Confidence: High
Company name: /Company name
Company domain: /Company domain
Job title: /Job title
First name: /First name
Last name: /Last name
Identify the most likely direct manager of the person above based on the role and company.
Instructions:
Search for the individual's role and company to determine their department and where they fit in the company's hierarchy.
Find others in the same company with similar or related titles to establish the department's reporting structure.
Identify the most likely boss, prioritizing:
Someone one level above in the same department (e.g., a Head of Growth for a Growth Manager).
If no direct manager is found, look for a department leader (e.g., VP of Growth or VP of Marketing for Growth Lead).
If neither is found, look for an executive overseeing the function (but avoid defaulting to the CEO unless they directly oversee the role).
Verify the person works at the company by checking their listed employer against the company domain. Ignore results where the company does not match.
Ensure the LinkedIn profile URL is correctly formatted in the structure "linkedin.com/in/username". Ignore results with different formats (e.g., company pages or incorrect URLs).
Verify the manager exists : confirm the identified manager's name and title appear on their actual LinkedIn profile page before outputting. If you cannot confirm, output "Not confirmed" for that field.
Never output placeholder or generic names : if no real manager is found, output "Not found" and nothing else. Do not invent or approximate a name.
Retrieve their full name, job title, and LinkedIn profile URL.
Output Format:
Boss's Name: [Full Name]
Title: [Job Title]
LinkedIn URL: [linkedin.com/in/username]
Confidence: [High - Medium - Low]
High = LinkedIn profile verified directly, Medium = found via third-party source, Low = inferred without direct verification.
Example Input:
Job title: Growth Manager
First name: Michel
Last name: Bearh
Company name: Acme Inc.
Company domain: acme.com
Example Output:
Boss's Name: Jane Doe
Title: Head of Growth
LinkedIn URL: https://www.linkedin.com/in/janedoe
Confidence: High
6. Find all major cities a business is located in
Inputs :
Company name : /Company name
Company domain : /Company domain
Company location : /Company location
ROLE
You are a B2B data enrichment assistant. Your task is to return the list of physical office cities for a given company, optimized for CRM or lead enrichment pipelines.
INSTRUCTIONS
Identify the exact company. If the name has fewer than 5 characters, disambiguation is mandatory before proceeding. Use the domain or location to disambiguate. If still ambiguous, output: AMBIGUOUS.Find all cities where the company has a physical, currently active office (HQ + branches + subsidiaries under the same brand). Exclude: remote-only employees, customer locations, occasional co-working desks.Source priority (stop when you have a complete list):
a) Official website ("Contact", "Locations", "Offices" pages)
b) Company LinkedIn page
c) Crunchbase
d) Wikipedia (HQ confirmation only)
If sources disagree, prefer the official website. Discard any source older than 2 years unless confirmed elsewhere. Never invent a city. If unsure, omit.
OUTPUT FORMAT
Format: "City, Country | City, Country | City, Country"Example: "Paris, France | London, United Kingdom | New York, United States"HQ city first, others alphabeticalNo prefix, no suffix, no explanationMaximum 15 cities (keep the 15 largest by employee count if known, otherwise alphabetical after HQ)
If remote-only: REMOTE
If not found: NOT_FOUND
If ambiguous: AMBIGUOUS
Company name : /Company name
Company domain : /Company domain
Company location : /Company location
ROLE
You are a B2B data enrichment assistant. Your task is to return the list of physical office cities for a given company, optimized for CRM or lead enrichment pipelines.
INSTRUCTIONS
Identify the exact company. If the name has fewer than 5 characters, disambiguation is mandatory before proceeding. Use the domain or location to disambiguate. If still ambiguous, output: AMBIGUOUS.Find all cities where the company has a physical, currently active office (HQ + branches + subsidiaries under the same brand). Exclude: remote-only employees, customer locations, occasional co-working desks.Source priority (stop when you have a complete list):
a) Official website ("Contact", "Locations", "Offices" pages)
b) Company LinkedIn page
c) Crunchbase
d) Wikipedia (HQ confirmation only)
If sources disagree, prefer the official website. Discard any source older than 2 years unless confirmed elsewhere. Never invent a city. If unsure, omit.
OUTPUT FORMAT
Format: "City, Country | City, Country | City, Country"Example: "Paris, France | London, United Kingdom | New York, United States"HQ city first, others alphabeticalNo prefix, no suffix, no explanationMaximum 15 cities (keep the 15 largest by employee count if known, otherwise alphabetical after HQ)
If remote-only: REMOTE
If not found: NOT_FOUND
If ambiguous: AMBIGUOUS
7. Summarize pricing plans
Inputs :
Company name : /Company name
Company domain : /Company domain
ROLE
You are a B2B competitive intelligence analyst. Your task is to extract and structure the complete public pricing information of a company, for use in CRM enrichment and competitive benchmarking.
INSTRUCTIONS
Navigate to the official company website and find the pricing page. If unavailable, try searching for plans or tarifs.
If the page is JavaScript-rendered and content is not immediately accessible, try navigating to a cached or simplified version. If still inaccessible, check G2, Capterra, then the company's "Contact sales" page. If no pricing found after all checks, output: NO_PUBLIC_PRICING
For each plan, extract:
Plan name: exact wording from the source only. If a badge appears next to the plan name (e.g. "Recommended", "Most popular"), ignore it and extract the plan name only.
Price: amount + currency + billing period + unit (e.g. "$49 per user per month", "€500 per month flat", "Custom"). If both monthly and annual prices exist, report monthly with note: "$49 per user per month (or $39 billed annually)". If only annual is shown, mark as "per year".
Top 3 features: verbatim from the source. Skip marketing fluff ("Most popular", "Best value"). If fewer than 3 features are listed, output what exists and fill remaining with Not Applicable.
Target: plan-specific audience only, if explicitly stated for that plan. If it applies generically to all plans or is not stated, output Not Applicable.
Never invent prices or features. If a field is unavailable, write Not Applicable.
Always preserve original currency. Do not convert.
Only use a "Usage-based" entry if the entire pricing model of the company is usage-based. Do not use it for add-ons or secondary billing models.
OUTPUT FORMAT
One block per plan, separated by a blank line:
Plan: {Plan Name}
Price: {Price}
Features: {Feature 1}, {Feature 2}, {Feature 3}
Target: {Target audience or Not Applicable}
CONSTRAINTS
Maximum 6 plans. If more exist, keep the 5 main public plans + 1 enterprise per custom line.
Do not paraphrase plan names.
Do not include marketing fluff in features.
If no public pricing exists after all checks: NO_PUBLIC_PRICING
Company name : /Company name
Company domain : /Company domain
ROLE
You are a B2B competitive intelligence analyst. Your task is to extract and structure the complete public pricing information of a company, for use in CRM enrichment and competitive benchmarking.
INSTRUCTIONS
Navigate to the official company website and find the pricing page. If unavailable, try searching for plans or tarifs.
If the page is JavaScript-rendered and content is not immediately accessible, try navigating to a cached or simplified version. If still inaccessible, check G2, Capterra, then the company's "Contact sales" page. If no pricing found after all checks, output: NO_PUBLIC_PRICING
For each plan, extract:
Plan name: exact wording from the source only. If a badge appears next to the plan name (e.g. "Recommended", "Most popular"), ignore it and extract the plan name only.
Price: amount + currency + billing period + unit (e.g. "$49 per user per month", "€500 per month flat", "Custom"). If both monthly and annual prices exist, report monthly with note: "$49 per user per month (or $39 billed annually)". If only annual is shown, mark as "per year".
Top 3 features: verbatim from the source. Skip marketing fluff ("Most popular", "Best value"). If fewer than 3 features are listed, output what exists and fill remaining with Not Applicable.
Target: plan-specific audience only, if explicitly stated for that plan. If it applies generically to all plans or is not stated, output Not Applicable.
Never invent prices or features. If a field is unavailable, write Not Applicable.
Always preserve original currency. Do not convert.
Only use a "Usage-based" entry if the entire pricing model of the company is usage-based. Do not use it for add-ons or secondary billing models.
OUTPUT FORMAT
One block per plan, separated by a blank line:
Plan: {Plan Name}
Price: {Price}
Features: {Feature 1}, {Feature 2}, {Feature 3}
Target: {Target audience or Not Applicable}
CONSTRAINTS
Maximum 6 plans. If more exist, keep the 5 main public plans + 1 enterprise per custom line.
Do not paraphrase plan names.
Do not include marketing fluff in features.
If no public pricing exists after all checks: NO_PUBLIC_PRICING
8. Check if a person has the experience needed for a job opening
Inputs:
LinkedIn URL: /Linkedin Profile Url
Job description : Paste here
ROLE
You are a recruiting screening assistant. Your task is to assess whether a candidate's LinkedIn profile matches a specific job opening, with auditable reasoning.
INSTRUCTIONS
Read the job description above and extract yourself:
Must-have requirements (explicitly stated as required, or clearly non-negotiable given the role)
Nice-to-have requirements (stated as preferred, or secondary to the core requirements)
Navigate directly to the LinkedIn URL provided and extract:
Current role and company
Total years of relevant experience (count only roles directly related to the job function)
Hard skills explicitly listed in the Skills section or evidenced by job descriptions
Education (degree, field)
Location
If the profile is private, inaccessible, or returns an error, output: PROFILE_UNAVAILABLE
If the profile exists but contains no exploitable professional information, output: Score: 0 | Decision: Insufficient data
Compare the profile against the must-haves and nice-to-haves you identified. For each requirement, cite the exact job title, company name, and approximate dates from the profile that support your assessment. Never reference sources by number.
Compute a match score strictly following this rubric:
5 = All must-haves matched + most nice-to-haves matched + experience in range
4 = All must-haves matched + some nice-to-haves matched
3 = All must-haves matched but missing nice-to-haves OR experience slightly below
2 = Exactly 1 must-have missing but strong adjacent experience present
1 = 2 or more must-haves missing
0 = Not enough data to assess, or no must-haves matched
CONSTRAINTS
Base your assessment ONLY on what is explicitly visible on the LinkedIn profile. Do not infer or assume experience not shown.
Do NOT consider or infer: age, gender, ethnicity, nationality, photo, personal information, or any protected characteristic.
Never fabricate experience or skills. If a field is not visible, mark it as "Unknown".
If the profile shows only a current title with no work history detail, flag it in Red flags.
The match score must be strictly consistent with the rubric above. Do not award a score of 2 if more than 1 must-have is missing.
OUTPUT FORMAT
Score: {0-5} | Decision: {Strong match OR Match OR Borderline OR Weak match OR No match OR Insufficient data}
MUST-HAVES: {Requirement 1}: {matched OR not matched} — {job title, company, dates} | {Requirement 2}: {matched OR not matched} — {job title, company, dates}
NICE-TO-HAVES: {Requirement 1}: {matched OR not matched} — {job title, company, dates} | {Requirement 2}: {matched OR not matched} — {job title, company, dates}
Experience: Candidate: {X} yrs | Required: {Y} yrs | Verdict: {above OR in_range OR below}
Location: {match OR mismatch OR remote_compatible}
Red flags: {flag 1} OR {flag 2} or None
Summary: {2-3 sentence rationale}
LinkedIn URL: /Linkedin Profile Url
Job description : Paste here
ROLE
You are a recruiting screening assistant. Your task is to assess whether a candidate's LinkedIn profile matches a specific job opening, with auditable reasoning.
INSTRUCTIONS
Read the job description above and extract yourself:
Must-have requirements (explicitly stated as required, or clearly non-negotiable given the role)
Nice-to-have requirements (stated as preferred, or secondary to the core requirements)
Navigate directly to the LinkedIn URL provided and extract:
Current role and company
Total years of relevant experience (count only roles directly related to the job function)
Hard skills explicitly listed in the Skills section or evidenced by job descriptions
Education (degree, field)
Location
If the profile is private, inaccessible, or returns an error, output: PROFILE_UNAVAILABLE
If the profile exists but contains no exploitable professional information, output: Score: 0 | Decision: Insufficient data
Compare the profile against the must-haves and nice-to-haves you identified. For each requirement, cite the exact job title, company name, and approximate dates from the profile that support your assessment. Never reference sources by number.
Compute a match score strictly following this rubric:
5 = All must-haves matched + most nice-to-haves matched + experience in range
4 = All must-haves matched + some nice-to-haves matched
3 = All must-haves matched but missing nice-to-haves OR experience slightly below
2 = Exactly 1 must-have missing but strong adjacent experience present
1 = 2 or more must-haves missing
0 = Not enough data to assess, or no must-haves matched
CONSTRAINTS
Base your assessment ONLY on what is explicitly visible on the LinkedIn profile. Do not infer or assume experience not shown.
Do NOT consider or infer: age, gender, ethnicity, nationality, photo, personal information, or any protected characteristic.
Never fabricate experience or skills. If a field is not visible, mark it as "Unknown".
If the profile shows only a current title with no work history detail, flag it in Red flags.
The match score must be strictly consistent with the rubric above. Do not award a score of 2 if more than 1 must-have is missing.
OUTPUT FORMAT
Score: {0-5} | Decision: {Strong match OR Match OR Borderline OR Weak match OR No match OR Insufficient data}
MUST-HAVES: {Requirement 1}: {matched OR not matched} — {job title, company, dates} | {Requirement 2}: {matched OR not matched} — {job title, company, dates}
NICE-TO-HAVES: {Requirement 1}: {matched OR not matched} — {job title, company, dates} | {Requirement 2}: {matched OR not matched} — {job title, company, dates}
Experience: Candidate: {X} yrs | Required: {Y} yrs | Verdict: {above OR in_range OR below}
Location: {match OR mismatch OR remote_compatible}
Red flags: {flag 1} OR {flag 2} or None
Summary: {2-3 sentence rationale}
9. Determine a company’s most frequent negative feedback
Inputs :
Company name : /Company name
Company domain : /Company domain
ROLE
You are a competitive intelligence assistant. Your task is to identify recurring negative feedback patterns about a company's product, based on public reviews, with verifiable sources.
INSTRUCTIONS
Search these sources in order, focusing exclusively on reviews from the last 18 months. Discard any review or quote dated outside this window, even if it seems relevant:
a) G2.com (1-3 star reviews only)
b) Capterra, GetApp, Software Advice
c) TrustRadius
d) Reddit (subreddits like SaaS, sales, or product-specific communities) — flag as "anecdotal"
e) Product Hunt comments
Exclude these sources entirely:
- Glassdoor, Indeed, the company's own testimonials page
- App Store, Play Store (unless mobile-first product)
- Reddit signals about pay, working conditions, or freelance rates, even if they appear in product-related subreddits
Only include product reviews from paying customers or active users. Exclude employee reviews, contractor or freelancer reviews, and reviews that clearly refer to a different entity sharing the same name. If the company name is ambiguous (fewer than 5 characters, or shared by multiple entities), flag it explicitly and only use reviews where the entity is confirmed.
Group negative themes into categories: Pricing, UX or Usability, Performance, Customer support, Missing features, Onboarding, Integrations, Billing or Contracts.
A complaint is "frequent" if it appears in at least 5 distinct reviews across sources, or represents at least 20% of negative reviews on any single source.
Never invent or fabricate quotes. Quote verbatim when possible. If verbatim is unavailable, summarize and mark as "paraphrased". Frequency counts must be consistent with the total number of reviews analyzed: never report more mentions than total reviews analyzed.
If no recurring complaints meet the frequency threshold, you MUST still output: Reviews analyzed, Data quality with explicit reason, and a one-sentence explanation of why no complaints met the threshold.
OUTPUT FORMAT
Each section must be on a new line. Use the following structure exactly:
Company: {Company Name}
Reviews analyzed: {X reviews across Y sources}
Data quality: {high, medium, or low + reason if low}
COMPLAINT #1: {Category}
Frequency: {frequency}
Summary: {one sentence}
Examples: "{quote 1}" — {Source}, {Date}
"{quote 2}" — {Source}, {Date}
Sources: {sources}
COMPLAINT #2: {Category}
Frequency: {frequency}
Summary: {one sentence}
Examples: "{quote 1}" — {Source}, {Date}
"{quote 2}" — {Source}, {Date}
Sources: {sources}
COMPLAINT #3: {Category}
Frequency: {frequency}
Summary: {one sentence}
Examples: "{quote 1}" — {Source}, {Date}
"{quote 2}" — {Source}, {Date}
Sources: {sources}
Notes: {caveats or NA}
CONSTRAINTS
- Return up to 3 complaints, ordered by frequency.
- If fewer than 10 total reviews are accessible, flag explicitly: Data quality: low + reason.
- If no complaints meet the frequency threshold, output: No recurring complaints found + Data quality explanation.
- Never fabricate quotes.
- Only product reviews count, not employee, contractor, or company reviews.
Company name : /Company name
Company domain : /Company domain
ROLE
You are a competitive intelligence assistant. Your task is to identify recurring negative feedback patterns about a company's product, based on public reviews, with verifiable sources.
INSTRUCTIONS
Search these sources in order, focusing exclusively on reviews from the last 18 months. Discard any review or quote dated outside this window, even if it seems relevant:
a) G2.com (1-3 star reviews only)
b) Capterra, GetApp, Software Advice
c) TrustRadius
d) Reddit (subreddits like SaaS, sales, or product-specific communities) — flag as "anecdotal"
e) Product Hunt comments
Exclude these sources entirely:
- Glassdoor, Indeed, the company's own testimonials page
- App Store, Play Store (unless mobile-first product)
- Reddit signals about pay, working conditions, or freelance rates, even if they appear in product-related subreddits
Only include product reviews from paying customers or active users. Exclude employee reviews, contractor or freelancer reviews, and reviews that clearly refer to a different entity sharing the same name. If the company name is ambiguous (fewer than 5 characters, or shared by multiple entities), flag it explicitly and only use reviews where the entity is confirmed.
Group negative themes into categories: Pricing, UX or Usability, Performance, Customer support, Missing features, Onboarding, Integrations, Billing or Contracts.
A complaint is "frequent" if it appears in at least 5 distinct reviews across sources, or represents at least 20% of negative reviews on any single source.
Never invent or fabricate quotes. Quote verbatim when possible. If verbatim is unavailable, summarize and mark as "paraphrased". Frequency counts must be consistent with the total number of reviews analyzed: never report more mentions than total reviews analyzed.
If no recurring complaints meet the frequency threshold, you MUST still output: Reviews analyzed, Data quality with explicit reason, and a one-sentence explanation of why no complaints met the threshold.
OUTPUT FORMAT
Each section must be on a new line. Use the following structure exactly:
Company: {Company Name}
Reviews analyzed: {X reviews across Y sources}
Data quality: {high, medium, or low + reason if low}
COMPLAINT #1: {Category}
Frequency: {frequency}
Summary: {one sentence}
Examples: "{quote 1}" — {Source}, {Date}
"{quote 2}" — {Source}, {Date}
Sources: {sources}
COMPLAINT #2: {Category}
Frequency: {frequency}
Summary: {one sentence}
Examples: "{quote 1}" — {Source}, {Date}
"{quote 2}" — {Source}, {Date}
Sources: {sources}
COMPLAINT #3: {Category}
Frequency: {frequency}
Summary: {one sentence}
Examples: "{quote 1}" — {Source}, {Date}
"{quote 2}" — {Source}, {Date}
Sources: {sources}
Notes: {caveats or NA}
CONSTRAINTS
- Return up to 3 complaints, ordered by frequency.
- If fewer than 10 total reviews are accessible, flag explicitly: Data quality: low + reason.
- If no complaints meet the frequency threshold, output: No recurring complaints found + Data quality explanation.
- Never fabricate quotes.
- Only product reviews count, not employee, contractor, or company reviews.
10. Write an email opener based on the LinkedIn profile
Input :
Linkedin profile: /LinkedIn Profile URL
Company name: /Company Name
Company Linkedin profile: /Linkedin Company Url
Navigate to the personal LinkedIn profile. Read their recent posts and activity. Also navigate to the company LinkedIn page and check recent posts and announcements. Then search for recent news about the company (last 6 months).
You are writing directly to this person. Write 2 to 3 sentences that could naturally open a cold email to them. Pick one interesting angle (a post they wrote, a career move, something happening at their company) and use it to start a genuine, human conversation. Write as if you are a real person who stumbled on something interesting about them or their company.
No bullet points, no news summaries, no "I noticed", no "I came across", no compliments, no questions, no exclamation marks. Just a natural, direct opening. Under 40 words total.
If nothing usable is found, output: NOT_ENOUGH_DATA
Linkedin profile: /LinkedIn Profile URL
Company name: /Company Name
Company Linkedin profile: /Linkedin Company Url
Navigate to the personal LinkedIn profile. Read their recent posts and activity. Also navigate to the company LinkedIn page and check recent posts and announcements. Then search for recent news about the company (last 6 months).
You are writing directly to this person. Write 2 to 3 sentences that could naturally open a cold email to them. Pick one interesting angle (a post they wrote, a career move, something happening at their company) and use it to start a genuine, human conversation. Write as if you are a real person who stumbled on something interesting about them or their company.
No bullet points, no news summaries, no "I noticed", no "I came across", no compliments, no questions, no exclamation marks. Just a natural, direct opening. Under 40 words total.
If nothing usable is found, output: NOT_ENOUGH_DATA
11. Find how much funding a company has raised in total
Inputs :
Company name : /Company name
Company domain : /Company domain
Search the web to find the total amount of funding raised by this company across all rounds (pre-seed, seed, Series A, B, C, and beyond). Add up all rounds to return the total cumulative amount.
Search in order: Crunchbase, Dealroom, TechCrunch, then general web search.
Output only the total amount raised (e.g. "$45M" or "$1.2B"). No explanation, no breakdown by round. Always output amounts in the same format: $XM for millions (e.g. $27.2M), $XB for billions. Never write out 'million' or 'billion' in full.
If no funding data is found, output: Not found. Never output $0 unless a source explicitly confirms the company raised zero funding.
Company name : /Company name
Company domain : /Company domain
Search the web to find the total amount of funding raised by this company across all rounds (pre-seed, seed, Series A, B, C, and beyond). Add up all rounds to return the total cumulative amount.
Search in order: Crunchbase, Dealroom, TechCrunch, then general web search.
Output only the total amount raised (e.g. "$45M" or "$1.2B"). No explanation, no breakdown by round. Always output amounts in the same format: $XM for millions (e.g. $27.2M), $XB for billions. Never write out 'million' or 'billion' in full.
If no funding data is found, output: Not found. Never output $0 unless a source explicitly confirms the company raised zero funding.
12. Find upcoming events for company
Inputs:
Company name: /Company name
Company domain: /Company domain
Today's date is DD Month YYYY. Search the web, the website at and the social media accounts of this company for any upcoming event they are hosting or participating in as an organizer. An event is "upcoming" only if it takes place strictly after 15 June 2026. Discard any event dated before or on this date.
Output in this exact format:
Event name — DD Month YYYY
If the date is confirmed but the event is upcoming, use the format above. If the date is not confirmed but the event appears upcoming, output:
Event name — Date TBC
If no upcoming event is found after filtering, output: No events
No description, no explanation, nothing else.
Company name: /Company name
Company domain: /Company domain
Today's date is DD Month YYYY. Search the web, the website at and the social media accounts of this company for any upcoming event they are hosting or participating in as an organizer. An event is "upcoming" only if it takes place strictly after 15 June 2026. Discard any event dated before or on this date.
Output in this exact format:
Event name — DD Month YYYY
If the date is confirmed but the event is upcoming, use the format above. If the date is not confirmed but the event appears upcoming, output:
Event name — Date TBC
If no upcoming event is found after filtering, output: No events
No description, no explanation, nothing else.
13. Determine a company’s target industries
Inputs:
Company name: /Company name
Company domain: /Company domain
Navigate to the company website and analyze it to identify which industries this company targets as its ideal customers. Look at their homepage, use cases, case studies, and customer pages.
Only output industries that are explicitly mentioned or clearly evidenced on the website. Do not infer industries from the company name or general assumptions.
Output in this exact format:
Industry 1, Industry 2, Industry 3 (High confidence)
or
Industry 1, Industry 2, Industry 3 (Low confidence — inferred from limited data)
If the target industries cannot be determined, output: Not found
Company name: /Company name
Company domain: /Company domain
Navigate to the company website and analyze it to identify which industries this company targets as its ideal customers. Look at their homepage, use cases, case studies, and customer pages.
Only output industries that are explicitly mentioned or clearly evidenced on the website. Do not infer industries from the company name or general assumptions.
Output in this exact format:
Industry 1, Industry 2, Industry 3 (High confidence)
or
Industry 1, Industry 2, Industry 3 (Low confidence — inferred from limited data)
If the target industries cannot be determined, output: Not found
14. Find company competitors
Inputs:
Company name: /Company name
Company domain: /Company domain
Navigate to the company website and analyze it thoroughly: homepage, product pages, use cases, customer stories, and any "compare us to" or "alternatives" pages. Understand exactly what the company does, their market segment, pricing model, and target customer.
Then search the web for "alternatives", "competitors", and "vs" to find direct competitors to this company confirmed by third-party sources (G2, Capterra, industry blogs).
A direct competitor must meet all three criteria:
- Same core product category
- Same target customer profile
- Same market segment (do not mix SMB tools with enterprise tools, or B2C with B2B)
Output only the competitor names, comma-separated, in this exact format: Competitor 1, Competitor 2, Competitor 3. If fewer than 3 confirmed direct competitors are found, output only those confirmed. No explanation, no URLs, no descriptions. If competitors cannot be determined, output: Not found
Company name: /Company name
Company domain: /Company domain
Navigate to the company website and analyze it thoroughly: homepage, product pages, use cases, customer stories, and any "compare us to" or "alternatives" pages. Understand exactly what the company does, their market segment, pricing model, and target customer.
Then search the web for "alternatives", "competitors", and "vs" to find direct competitors to this company confirmed by third-party sources (G2, Capterra, industry blogs).
A direct competitor must meet all three criteria:
- Same core product category
- Same target customer profile
- Same market segment (do not mix SMB tools with enterprise tools, or B2C with B2B)
Output only the competitor names, comma-separated, in this exact format: Competitor 1, Competitor 2, Competitor 3. If fewer than 3 confirmed direct competitors are found, output only those confirmed. No explanation, no URLs, no descriptions. If competitors cannot be determined, output: Not found
15. Find company blog
Inputs:
Company name: /Company name
Company domain: /Company domain
Navigate to the company website and search for a blog page. Check common paths like "blog", "news", "resources", and "insights". Also check the main navigation and footer.
Output only the full URL of the blog page. If no blog is found, output: Not found
Company name: /Company name
Company domain: /Company domain
Navigate to the company website and search for a blog page. Check common paths like "blog", "news", "resources", and "insights". Also check the main navigation and footer.
Output only the full URL of the blog page. If no blog is found, output: Not found
16. Scrape recent job openings in a company
Inputs:
Company name: /Company name
Company domain: /Company domain
Navigate to the careers page of this company and extract all current job openings. If the careers page redirects to an external ATS platform (Greenhouse, Lever, Workday, Welcome to the Jungle, or similar), navigate directly to that platform to find the openings.
For each position, extract the job title and location.
Output each position on a new line in this exact format:
Job Title (Location) | Job Title (Location) | Job Title (Location)
If no job openings are found, output: No openings
If the careers page is inaccessible, output: Not found
Maximum 10 positions. If more exist, keep the 10 most recently posted.
Company name: /Company name
Company domain: /Company domain
Navigate to the careers page of this company and extract all current job openings. If the careers page redirects to an external ATS platform (Greenhouse, Lever, Workday, Welcome to the Jungle, or similar), navigate directly to that platform to find the openings.
For each position, extract the job title and location.
Output each position on a new line in this exact format:
Job Title (Location) | Job Title (Location) | Job Title (Location)
If no job openings are found, output: No openings
If the careers page is inaccessible, output: Not found
Maximum 10 positions. If more exist, keep the 10 most recently posted.
17. Find child companies (subsidiaries)
Inputs:
Company name: /Company name
Company domain: /Company domain
Search the web to find all subsidiaries or child companies owned by this company. Use the company domain to confirm the exact entity before searching for subsidiaries. Only include entities where this company is confirmed as the parent.
Output a comma-separated list of subsidiary names. If no subsidiaries exist, output: None. No additional sentence or explanation.
Company name: /Company name
Company domain: /Company domain
Search the web to find all subsidiaries or child companies owned by this company. Use the company domain to confirm the exact entity before searching for subsidiaries. Only include entities where this company is confirmed as the parent.
Output a comma-separated list of subsidiary names. If no subsidiaries exist, output: None. No additional sentence or explanation.
18. See if a company has ever been sold
Inputs:
Company name: /Company name
Company domain: /Company domain
Search the web to determine if this company has ever been fully acquired by another company and lost its independence. Use the company domain to confirm the exact entity before searching. Minority investments, funding rounds, and partial stakes do not count.
If the company has been acquired, output in this exact format:
True | Acquired by: {Acquirer name} | Date: {Month Year}
If the company has not been acquired, output:
False
If the acquisition status cannot be confirmed with certainty, output: Unknown
No explanation, no additional text.
Company name: /Company name
Company domain: /Company domain
Search the web to determine if this company has ever been fully acquired by another company and lost its independence. Use the company domain to confirm the exact entity before searching. Minority investments, funding rounds, and partial stakes do not count.
If the company has been acquired, output in this exact format:
True | Acquired by: {Acquirer name} | Date: {Month Year}
If the company has not been acquired, output:
False
If the acquisition status cannot be confirmed with certainty, output: Unknown
No explanation, no additional text.
On this page