Complete step-by-step operational playbook for Leo Kuyumciyan (Phellos Consultancy). Two tracks - startup founders for podcast invites and VC/PE contacts for referral outreach. From Sales Navigator filters to tiered Instantly + HeyReach campaigns with dossier-based personalization.
Build two lead pipelines - startup founders (podcast invite) and VC/PE investors (referral intro) - from raw Sales Navigator searches through enrichment, verification, scoring, and research into campaign-ready CSVs for Instantly and HeyReach.
Download the Claude Code skill file for this pipeline. Load it into any Claude Code session to get an AI co-pilot that knows every step, catches deviations from the SOP, and suggests the best approach at each decision point.
The skill file includes: full pipeline reference, scoring rubrics, DNC list, deviation detection (16 common mistakes it flags automatically), FAQ with strategic rationale for every decision, and tool budget tracking.
How to use: Open terminal, type claude, then drag the downloaded .md file into the Claude Code window. It loads automatically.
Tool access required + estimated pipeline time
Make sure you have active access to ALL of these before starting Step 1:
| Tool | Used In | Cost | New? |
|---|---|---|---|
| LinkedIn Sales Navigator | Step 1 | Existing subscription | No |
| HeyReach | Steps 2, 8 | Existing account | No |
| Crunchbase | Step 2B | Free (basic filters) | No |
| Crunchbase Pro | Funding Triggers (ongoing) | $49/mo - optional (see Funding Triggers section) | New if not subscribed |
| Instant Data Scraper | Step 2B | Free Chrome extension | No |
| Prospeo | Step 4A | Existing (~2K credits/month) | No |
| Apollo | Step 4B | $109/mo, 200 reveals/day cap - UK segments only | No |
| Kaspr | Step 4C | EUR99/mo, 250 reveals/month - UK segments 1, 2, 5 only | NEW |
| MillionVerifier | Step 5 | Pay-per-credit (~$0.003/email, ~$15-20 per pipeline run) | No |
| Claude Code | Steps 3, 4A, 4C, 5, 6, 7, 7B, 8 | Existing Anthropic subscription | No |
| Instantly | Step 8 | Existing (18 active accounts) | No |
API keys for Prospeo and Apollo must be in ~/.env as PROSPEO_API_KEY and APOLLO_API_KEY before running Claude Code prompts.
Login credentials: All tool logins are in the shared 1Password vault under "Phellos Lead Gen." If you don't have 1Password access, ask Jordan before starting. Claude Code: open your Mac terminal, type claude and press Enter. That's all - no install needed.
This pipeline runs across 1-2 days. Most waits are tool-processing time, not active work.
| Step | Active Work | Wait Time |
|---|---|---|
| Step 1: Build Sales Nav searches | 2-3 hours | - |
| Step 2: HeyReach scraping | 30 min setup | 1-3 hours (runs in background) |
| Step 2B: Crunchbase | 30 min | - |
| Step 3: Claude Code dedup/filter | 15 min (paste prompts) | 2-5 min per prompt |
| Step 4A: Prospeo enrichment | 5 min (paste prompt) | 30-60 min (API batches) |
| Step 4B: Apollo backfill | 20 min (manual upload) | - |
| Step 4C: Kaspr UK recovery | 15 min (manual upload) | - |
| Step 4D: Recombine | 5 min | 2-5 min |
| Step 5: MillionVerifier | 10 min (upload) | 5-30 min (runs in background) |
| Step 6: Score + Tier | 5 min | 2-5 min |
| Step 7: Sherlock research | 5 min (paste prompt) | ~2.5 hours (3 min/lead x 50) |
| Step 7B: Hook generation | 5 min | 10-20 min |
| Step 8: Build CSVs + QA | 20 min | 5-10 min |
| Total | ~5 hours active | ~5-7 hours background |
Day 1: Steps 1-2B (building searches and scraping). Day 2: Steps 3-8 (all Claude Code prompts, enrichment, and CSV build). Sherlock (Step 7) can run overnight.
This playbook incorporates 7 improvements discovered during the Jamie lead pipeline build:
build_campaigns.py handles all tiers and tracks.Based on Leo's Dec 23, 2025 call feedback and Feb 25, 2026 VC/PE handoff.
| Filter | Value |
|---|---|
| Company Age | 2-3 years (max 5-6 unless mid-round) |
| Employees | 10-30 sweet spot (11-50 in Sales Nav) |
| Sector | SaaS, Fintech, hardware-enabled tech |
| Geography | UK + USA |
| Decision Maker | CEO / Founder |
| Revenue | Under $40M |
| Trigger | Actively fundraising or raised in last 12mo |
| ICP Segment | Est. Leads |
|---|---|
| Venture Capital + Private Equity | ~8,000 |
| Investment Management | ~6,500 |
| Capital Markets | ~8,000 |
| M&A Advisory | ~3,500 |
| Partners, GPs, MDs at UK + US funds investing in SaaS/Fintech/hardware startups | |
HeyReach scraping is free. Qualify before enriching to avoid wasting credits. Leo's ICP is tighter than Jamie's - company age (2-3 years) and funding stage matter more than just title + headcount.
All lead source URLs are tracked in one master Google Sheet. Every URL you build in Step 1 must be saved here before scraping. Open the "Phellos" tab, find the matching sub-search row, and paste the URL.
How to build each search (do this for every sub-search in the tables below):
Go to linkedin.com/sales/search/people (Sales Navigator Lead Search)
You will see a search bar at the top and filter panels on the left sidebar. DO NOT type anything into the top search bar - leave it blank always.
On the left sidebar, click "Company headcount". Check the box for the headcount range listed in the sub-search table (e.g., 11-50). Click "Show results".
Click "Current job title" in the sidebar. A text box appears. Type or paste the INCLUDE titles: CEO OR Founder OR Co-Founder OR Chief Executive. Press Enter. They will appear as green pills (green = included). Confirm the toggle at the top of the title filter says "Include."
Now add the EXCLUDE titles. Click the "Include" dropdown at the top of the title filter and switch it to "Exclude". Now type or paste: GTM OR Growth OR Advisor OR Board OR Consultant OR Coach OR Fractional OR Intern OR Associate. Press Enter. They will appear as red pills (excluded). Switch the dropdown back to "Include" when done. Alternatively, type values while on "Include" (green pills), then click the circle-with-line-through-it icon next to each to turn them red (excluded).
Click "Geography" in the sidebar. Type the geography from the segment table below and select it from the dropdown.
Click "Industry" in the sidebar. Type and select the INCLUDE industries from the segment table. They appear as green pills. Then add any EXCLUDE industries - type them, and click the circle-with-line-through-it icon to turn them red.
Click "Seniority level" in the sidebar. Check Owner only. Do NOT check CXO.
Click "Posted on LinkedIn" in the sidebar. If the sub-search says "ON," toggle to Yes. If "OFF," leave untouched.
CHECK THE RESULT COUNT in the top-right area. If it says "2,500+ results", you must split further. If under 2,500, you are good.
Copy the full URL from your browser's address bar. Open the Lead Source URL Tracker (Phellos tab), find the matching sub-search row, and paste the URL. Also enter the actual result count.
Repeat for each sub-search in the segment tables below.
Valid headcount options ONLY: Self-employed, 1-10, 11-50, 51-200, 201-500, 501-1000, 1001-5000, 5001-10000, 10001+. There are NO custom ranges (no 10-30, no 51-100).
Industry names that do NOT exist: Computer Software (use Software Development), IT and Services (use Technology, Information and Internet), Internet (merged into Technology, Information and Internet), Computer Networking (merged into Technology, Information and Internet).
2,500 lead limit: HeyReach cannot scrape more than 2,500 from a single search URL. If a search returns 2,500+, split it using: (1) Posted ON/OFF, (2) Headcount split, (3) Geography split.
Shared base filters for ALL Founders Track searches:
| Filter | Value | Where to Find It |
|---|---|---|
| Title INCLUDE | CEO OR Founder OR Co-Founder OR Chief Executive | Left sidebar > Current job title > Include |
| Title EXCLUDE | GTM OR Growth OR Advisor OR Board OR Consultant OR Coach OR Fractional OR Intern OR Associate | Left sidebar > Current job title > Exclude (red pills) |
| Seniority | Owner only | Left sidebar > Seniority level |
| Top keyword bar | LEAVE BLANK ALWAYS | Top of page - do not type here |
Leo's sweet spot. Early-stage UK SaaS companies (typically Seed to Series A) with 10-30 employees. Company is 2-3 years old, actively fundraising or recently raised. Board pressure to demonstrate traction.
| Filter | Value | How to Set |
|---|---|---|
| Company Headcount | 11-50 | Left sidebar > Company headcount > check "11-50" |
| Geography | United Kingdom | Left sidebar > Geography > type and select |
| Title INCLUDE | CEO OR Founder OR Co-Founder OR Chief Executive | Current job title > Include dropdown > paste |
| Title EXCLUDE | GTM OR Growth OR Advisor OR Board OR Consultant OR Coach OR Fractional OR Intern OR Associate | Current job title > switch to Exclude > paste (red pills) |
| Industry INCLUDE | Software Development + Technology, Information and Internet | Industry filter > type each one > green pills |
| Industry EXCLUDE | Financial Services | Industry filter > type > click stop-sign icon > red pill |
| Seniority | Owner | Seniority level > check Owner ONLY |
| Posted on LinkedIn | Varies per sub-search (see table below) | Left sidebar > Posted on LinkedIn |
| Top keyword bar | LEAVE BLANK | Do not type anything in the top search bar |
| # | Sub-Search | Posted | Est. Count | Sales Nav URL |
|---|---|---|---|---|
| 1a | SaaS UK Early Stage - Posted (11-50) | ON | ~800 | Build search |
| 1b | SaaS UK Early Stage - Not Posted (11-50) | OFF | ~1,200 | Build search |
~2,000 raw across 2 sub-searches. Heavy overlap between Posted/Not Posted is expected.
Leo's sweet spot is 10-30 employees (11-50 in Sales Nav). Companies with 51-200 employees typically have finance teams already and would hire FTEs instead of using Leo's services. Only scrape this segment after Segments 1, 3, 5, and 6 are fully processed and you need more volume.
Larger UK SaaS. Typically Series B companies with more team but growth per dollar invested is declining. Leo's threshold: under $40M revenue, and only relevant if actively mid-round or raised recently.
| Filter | Value | How to Set |
|---|---|---|
| Company Headcount | 51-200 | Left sidebar > Company headcount > check "51-200" |
| Geography | United Kingdom | Left sidebar > Geography > type and select |
| Title INCLUDE | CEO OR Founder OR Co-Founder OR Chief Executive | Current job title > Include dropdown > paste |
| Title EXCLUDE | GTM OR Growth OR Advisor OR Board OR Consultant OR Coach OR Fractional OR Intern OR Associate | Current job title > switch to Exclude > paste (red pills) |
| Industry INCLUDE | Software Development + Technology, Information and Internet | Industry filter > type each one > green pills |
| Industry EXCLUDE | Financial Services | Industry filter > type > click stop-sign icon > red pill |
| Seniority | Owner | Seniority level > check Owner ONLY |
| Posted on LinkedIn | Varies per sub-search | Left sidebar > Posted on LinkedIn |
| # | Sub-Search | Posted | Est. Count | Sales Nav URL |
|---|---|---|---|---|
| 2a | SaaS UK Growth Stage - Posted (51-200) | ON | ~600 | Build search |
| 2b | SaaS UK Growth Stage - Not Posted (51-200) | OFF | ~900 | Build search |
~1,500 raw across 2 sub-searches.
US market expansion. Same ICP as Segment 1 but US geography. Larger pool - will likely need more sub-search splits to stay under 2,500.
| Filter | Value | How to Set |
|---|---|---|
| Company Headcount | 11-50 | Left sidebar > Company headcount > check "11-50" |
| Geography | United States | Left sidebar > Geography > type and select |
| Title INCLUDE | CEO OR Founder OR Co-Founder OR Chief Executive | Current job title > Include dropdown > paste |
| Title EXCLUDE | GTM OR Growth OR Advisor OR Board OR Consultant OR Coach OR Fractional OR Intern OR Associate | Current job title > switch to Exclude > paste (red pills) |
| Industry INCLUDE | Software Development + Technology, Information and Internet | Industry filter > type each one > green pills |
| Industry EXCLUDE | Financial Services | Industry filter > type > click stop-sign icon > red pill |
| Seniority | Owner | Seniority level > check Owner ONLY |
| Posted on LinkedIn | Varies per sub-search | Left sidebar > Posted on LinkedIn |
| # | Sub-Search | Posted | Est. Count | Sales Nav URL |
|---|---|---|---|---|
| 3a | SaaS US Early Stage - Posted (11-50) | ON | ~1,500 | Build search |
| 3b | SaaS US Early Stage - Not Posted (11-50) | OFF | ~1,500 | Build search |
~3,000 raw across 2+ sub-searches. WILL need further splitting - do not scrape any sub-search showing 2,500+.
Leo's sweet spot is 10-30 employees (11-50 in Sales Nav). Companies with 51-200 employees typically have finance teams already and would hire FTEs instead of using Leo's services. Only scrape this segment after Segments 1, 3, 5, and 6 are fully processed and you need more volume.
Larger US SaaS. Same as Segment 2 but US geography. Only relevant if under $40M revenue and actively mid-round.
| Filter | Value | How to Set |
|---|---|---|
| Company Headcount | 51-200 | Left sidebar > Company headcount > check "51-200" |
| Geography | United States | Left sidebar > Geography > type and select |
| Title INCLUDE | CEO OR Founder OR Co-Founder OR Chief Executive | Current job title > Include dropdown > paste |
| Title EXCLUDE | GTM OR Growth OR Advisor OR Board OR Consultant OR Coach OR Fractional OR Intern OR Associate | Current job title > switch to Exclude > paste (red pills) |
| Industry INCLUDE | Software Development + Technology, Information and Internet | Industry filter > type each one > green pills |
| Industry EXCLUDE | Financial Services | Industry filter > type > click stop-sign icon > red pill |
| Seniority | Owner | Seniority level > check Owner ONLY |
| Posted on LinkedIn | Varies per sub-search | Left sidebar > Posted on LinkedIn |
| # | Sub-Search | Posted | Est. Count | Sales Nav URL |
|---|---|---|---|---|
| 4a | SaaS US Growth Stage - Posted (51-200) | ON | ~1,200 | Build search |
| 4b | SaaS US Growth Stage - Not Posted (51-200) | OFF | ~1,300 | Build search |
~2,500 raw across 2 sub-searches.
Fintech is Leo's second vertical. Note: this segment INCLUDES Financial Services (which was excluded in Segments 1-4) combined with Technology. This catches fintech founders specifically.
| Filter | Value | How to Set |
|---|---|---|
| Company Headcount | 11-50 | Left sidebar > Company headcount > check "11-50" |
| Geography | Varies per sub-search (UK or US) | Left sidebar > Geography > type and select |
| Title INCLUDE | CEO OR Founder OR Co-Founder OR Chief Executive | Current job title > Include dropdown > paste |
| Title EXCLUDE | GTM OR Growth OR Advisor OR Board OR Consultant OR Coach OR Fractional OR Intern OR Associate | Current job title > switch to Exclude > paste (red pills) |
| Industry INCLUDE | Financial Services + Technology, Information and Internet | Industry filter > type each one > green pills |
| Seniority | Owner | Seniority level > check Owner ONLY |
| Posted on LinkedIn | Varies per sub-search | Left sidebar > Posted on LinkedIn |
| # | Sub-Search | Geography | Posted | Est. Count | Sales Nav URL |
|---|---|---|---|---|---|
| 5a | Fintech UK - Posted (11-50) | United Kingdom | ON | ~600 | Build search |
| 5b | Fintech US - Posted (11-50) | United States | ON | ~800 | Build search |
| 5c | Fintech UK - Not Posted (11-50) | United Kingdom | OFF | ~600 | Build search |
~2,000 raw across 3 sub-searches.
Leo's dream examples include Remarkable (hardware-enabled). Companies building physical products with tech components - IoT, robotics, connected devices, hardware SaaS hybrids.
| Filter | Value | How to Set |
|---|---|---|
| Company Headcount | 11-50 | Left sidebar > Company headcount > check "11-50" |
| Geography | Varies per sub-search (UK or US) | Left sidebar > Geography > type and select |
| Title INCLUDE | CEO OR Founder OR Co-Founder OR Chief Executive | Current job title > Include dropdown > paste |
| Title EXCLUDE | GTM OR Growth OR Advisor OR Board OR Consultant OR Coach OR Fractional OR Intern OR Associate | Current job title > switch to Exclude > paste (red pills) |
| Industry INCLUDE | Technology, Information and Internet + Telecommunications | Industry filter > type each one > green pills |
| Industry EXCLUDE | Financial Services | Industry filter > type > click stop-sign icon > red pill |
| Seniority | Owner | Seniority level > check Owner ONLY |
| Posted on LinkedIn | Varies per sub-search | Left sidebar > Posted on LinkedIn |
| # | Sub-Search | Geography | Posted | Est. Count | Sales Nav URL |
|---|---|---|---|---|---|
| 6a | Hardware-Enabled UK (11-50) | United Kingdom | ON | ~400 | Build search |
| 6b | Hardware-Enabled US - Posted (11-50) | United States | ON | ~600 | Build search |
| 6c | Hardware-Enabled US - Not Posted (11-50) | United States | OFF | ~500 | Build search |
~1,500 raw across 3 sub-searches.
The VC/PE track was segmented in the Feb 25 handoff. Four ICPs are already defined with all filter values. They are tracked in the same Lead Source URL Tracker (Phellos tab).
| Filter | Value |
|---|---|
| Company Headcount | 1-10 + 11-50 (ICPs 1-3 only - see ICP 4 override below) |
| Company Type | Privately Held |
| Geography | United Kingdom + United States |
| Posted on LinkedIn | Last 30 days |
| Title INCLUDE | Partner OR General Partner OR Managing Partner OR Principal OR Managing Director OR Investment Director OR Head of Portfolio OR Portfolio Director |
| Title EXCLUDE | Advisor OR Board Member OR Consultant OR Fractional OR Intern OR Associate OR Analyst OR Operating Partner OR Limited Partner OR Venture Partner |
| ICP | Industry | Headcount Override | Est. Count |
|---|---|---|---|
| ICP 1: Venture Capital + Private Equity | Venture Capital and Private Equity Principals | Shared (1-10 + 11-50) | ~8,000 |
| ICP 2: Investment Management | Investment Management | Shared (1-10 + 11-50) | ~6,500 |
| ICP 3: Capital Markets | Capital Markets | Shared (1-10 + 11-50) | ~8,000 |
| ICP 4: M&A Advisory (Investment Banking) | Investment Banking | OVERRIDE: 51-200 + 201-500 + 501-1000 + 1001-5000 - investment banks are large firms | ~3,500 |
Do this for each sub-search URL you saved in Step 1:
Log in to app.heyreach.io. Make sure you are in the Phellos workspace (top-left dropdown). All Phellos lead lists and campaigns must be in the Phellos workspace, not the Dopamine Digital workspace.
In the left sidebar, click "Lead Lists"
Click the "+ Create New List" button (top right)
Name the list to match the sub-search. Example: Phellos - SaaS UK Early 11-50 Posted (1a)
Select "Import from Sales Navigator URL"
Paste the full Sales Nav URL from Step 1. Click "Import".
Scraping takes 5-15 minutes. You can start another sub-search while this one runs. The status column in your Lead Lists view will show "Processing" (with a spinning icon) while it runs, then change to "Completed" (green) when done. Refresh the page if the status hasn't updated after 20 minutes.
When status changes to "Completed", click the list name to open it.
Click "Export" (top right) > "Download as CSV".
Save the file as Phellos_Raw_[ID].csv (e.g., Phellos_Raw_1a.csv).
Repeat for all remaining sub-searches (1b, 2a, 2b, 3a, 3b, 4a, 4b, 5a, 5b, 5c, 6a, 6b, 6c, plus all VC sub-searches).
~/Downloads/Phellos-Lead-Gen/1-Raw-Scrapes/. Each import takes 5-15 minutes to process. Start all imports first, then come back to download when they show "Completed." You can start a new import while the previous one is still running.Purpose: Source recently-funded startups matching Leo's ICP. Crunchbase gives you companies with funding data (stage, amount, date). Combined with Sales Nav leads, this catches founders who might not show up in LinkedIn searches but have the right funding profile.
Part 1: Install Instant Data Scraper (one-time setup)
Open Google Chrome. Go to Instant Data Scraper on Chrome Web Store
Click "Add to Chrome" > "Add Extension". Free, no account needed.
Part 2: Set Crunchbase filters for Phellos ICP
Go to crunchbase.com/discover/organization.companies. Log in if prompted.
Overview card: Click the + icon. Set Industry: Software, Financial Technology, Information Technology. Set Headquarters Location: United Kingdom + United States. Set Number of Employees: drag slider to 11-100. Set Founded Date: Custom Date Range, 2020-2024 (catches Leo's 2-3 year old preference).
Financials card: Click the + icon. Set Last Funding Type: check Seed. (Run Series A and Series B as separate searches.) Set Last Funding Date: Past Year.
Company Status card: Click the + icon. Set Type: For Profit. Operating Status: Active. IPO Status: Private.
Save the search. Name it: Phellos software+fintech, 11-100 emp, Seed, 2020-2024, UK+US
| Filter Card | Filter Name | Value | UI Control |
|---|---|---|---|
| Overview | Industry | Software, Financial Technology, Information Technology | Search box > type > select pill |
| Overview | HQ Location | United Kingdom, United States | Search box > type > select pills |
| Overview | Employees | 11 to 100 | Drag slider handles |
| Overview | Founded Date | 2020-2024 | Custom Date Range |
| Financials | Last Funding Type | Seed (then Series A, then Series B - separate searches) | Checkbox |
| Financials | Last Funding Date | Past Year | Radio button |
| Company Status | Type | For Profit | Checkbox |
| Company Status | Operating Status | Active | Checkbox |
| Company Status | IPO Status | Private | Checkbox |
Part 3: Scrape with Instant Data Scraper
On the Crunchbase results page, click the Instant Data Scraper icon in your Chrome toolbar (puzzle-piece icon at top right of Chrome > find "Instant Data Scraper" and pin it). A popup appears with a table preview of the page data.
Click "Locate 'Next' button" - the tool will flash a blue highlight over different elements on the page. Keep clicking until it highlights the "Next" pagination arrow at the bottom right of the Crunchbase results. If no blue highlight appears after 3 clicks, scroll down to make the pagination visible first.
Set delay: Min: 2 sec, Max: 20 sec. Leave "Infinite scroll" unchecked.
Click "Start crawling". Let it run through all pages.
When done, click "CSV" to download. Save as Phellos_Crunchbase_Seed.csv
Repeat for Series A and Series B searches (change only the Last Funding Type checkbox).
~/Downloads/Phellos-Lead-Gen/2-Crunchbase-Sources/Why Dealroom: Dealroom has better UK early-stage coverage than Crunchbase - it catches sub-$5M rounds that Crunchbase misses. The free tier allows browsing and manual copy.
Go to app.dealroom.co. Create a free account if you don't have one.
Set filters: Location = United Kingdom, Sector = SaaS OR Fintech OR Hardware, Stage = Seed OR Series A, Founded = 2020-2024, Employees = 10-100.
Browse results. For each qualifying company, manually copy company name + website + founding year into a spreadsheet.
Save as Phellos_Dealroom_UK.csv in ~/Downloads/Phellos-Lead-Gen/2-Crunchbase-Sources/
Phellos_Dealroom_UK.csv with company name, website, founding year. Save to ~/Downloads/Phellos-Lead-Gen/2-Crunchbase-Sources/. These leads enter the pipeline at Step 3 Prompt 1 alongside the Crunchbase CSVs - drag this file in along with the others when running the Merge + Dedup prompt. The Prospeo enrichment in Step 4A cannot process them directly (no LinkedIn URL), so any Dealroom company that does NOT match a Founders Track lead by domain in Step 4D will be skipped for email campaigns but can be manually researched for T0 outreach.Why accelerator portfolios: Leo specifically mentioned Seedcamp-backed startups as a good fit. These portfolio pages are public and the companies are pre-qualified - they've been vetted by top-tier accelerators, have funding, and are in growth mode.
Go to seedcamp.com/our-companies/. Browse all companies.
Filter mentally for: SaaS/Fintech/hardware, UK-based or UK-connected, founded 2019-2024, under 100 employees. To tell SaaS from non-SaaS: SaaS companies sell software subscriptions (look for words like "platform," "software," "API," "dashboard," "analytics" in their description). Non-SaaS includes physical goods, consulting services, restaurants, and anything without a software product. If in doubt, copy the company and Jordan/Leo will review.
Copy company name + website + tag as "Seedcamp" into a CSV.
Repeat for techstars.com/portfolio (filter for London/UK programs).
Optionally add: joinef.com/portfolio/ (Entrepreneur First - London-based accelerator).
Save as Phellos_Accelerator_Portfolio.csv in ~/Downloads/Phellos-Lead-Gen/2-Crunchbase-Sources/
Tag all leads with accelerator_backed in the source column for scoring bonus.
Phellos_Accelerator_Portfolio.csv with company name, website, accelerator tag. Save to ~/Downloads/Phellos-Lead-Gen/2-Crunchbase-Sources/. These leads enter the pipeline at Step 3 Prompt 1 alongside the Crunchbase CSVs - drag this file in along with the others when running the Merge + Dedup prompt. The accelerator_backed source tag flows through to scoring in Step 6 where it earns +15 points. Leads that don't match a Founders Track lead by domain in Step 4D are skipped for email campaigns but can be manually researched for T0 outreach.In Jamie's v1, DNC filtering happened after enrichment - wasting credits on leads that would be removed anyway. This playbook filters Turkey companies and existing client DNC matches at Step 3, before spending any Prospeo or Apollo credits.
Run these 4 prompts in sequence in Claude Code. Open your Mac terminal, type claude and press Enter. You will see a > prompt. To "drag in" a file: open a Finder window, find the CSV file, then drag it from Finder and drop it directly into the terminal window - Claude Code will read it automatically. Then paste the prompt text below and press Enter. Wait for Claude Code to finish (it will print results) before running the next prompt.
When a person holds a senior role at a VC/PE firm AND also appears as a co-founder of a startup, they are two different opportunities - referral (via their VC hat) vs podcast guest (via their founder hat). The correct track depends on seniority:
dual_role=true and route to VC track as a default. VC outreach is less risky - a VC who also founded a company will not be confused by a referral ask, but a founder who is not actively investing will find a portfolio-resource email irrelevant.Before running the prompt: scan the cross-track duplicate list for rows where the person's VC title is GP / Managing Partner / Partner at a named fund. Manually move those rows to the VC track before running the merge. The prompt's default rule (keep in Founders) is correct for edge cases, but named-fund partners should go to VC.
Prospeo has better data for UK leads (Leo's primary market). Apollo fills gaps on US leads where Prospeo misses. This order maximizes hit rate and minimizes Apollo credit usage.
Prospeo is the primary enrichment tool. It uses bulk LinkedIn profile enrichment to find verified business emails. Expected hit rate: ~40-50%.
Apollo fills gaps on leads Prospeo missed. Expected hit rate on Prospeo misses: ~55-65%.
100 reveals per run. 200 reveals per day. NEVER bypass these limits without Jordan's explicit approval. Apollo credits cost real money and the pipeline burned 1,000 credits in 40 minutes once due to a missing cap.
Since scoring hasn't happened yet at this step, use a segment proxy: only upload leads from Segment 1-2 (UK SaaS - Leo's sweet spot) and Segment 5 (Fintech UK portion) to Apollo. Skip Segment 3-4 (US SaaS) and Segment 6 (Hardware) entirely - those Prospeo misses go straight to the LinkedIn-only file for HeyReach outreach.
This reduces Apollo credit usage by approximately 50% while keeping the highest-value UK leads fully enriched.
Apollo credit math: Segments 1+2+5 raw = ~5,500 leads. After dedup/filter: ~3,000. Prospeo hits ~45%: ~1,350 found, ~1,650 misses. Of those misses, UK-only (Segs 1+2+5): ~1,350 (US leads excluded). Apollo hit rate on UK misses: ~55% = ~750 reveals. At 200/day cap: ~4 days of reveals. Total Apollo spend: ~750 credits out of the 5K/month plan ($109/mo). Well within budget.
Go to app.apollo.io. Make sure you are in the Phellos Apollo account. Click People in the left sidebar.
Before uploading: Filter the Prospeo misses CSV to UK segments only (Segments 1, 2, 5). Use Claude Code: drag in Phellos_Founders_Prospeo_Misses.csv and run: "Filter this misses file to only keep leads where segment_id is 1, 2, or 5 (UK SaaS and Fintech segments). Save the filtered file as Phellos_Founders_Prospeo_Misses_UK.csv." Upload ONLY this filtered file to Apollo. US segments (3, 4) and Hardware (6) go directly to the LinkedIn-only file - skip Apollo for them.
Click Import > Upload CSV. Upload the UK-filtered misses file. Then separately upload VC misses.
Apollo will match leads and show available emails. Do NOT reveal more than 100 at a time. To reveal: select leads using the checkboxes, then click "Access emails & phone". Apollo shows a counter of how many reveals you are about to use. If the counter exceeds 100, deselect some leads before confirming. Each reveal costs 1 credit - there is no undo.
Export the results as CSV. Save as Phellos_Founders_Apollo.csv / Phellos_VC_Apollo.csv
~/Downloads/Phellos-Lead-Gen/4-Enrich/Why Kaspr: Kaspr has better EU/UK email data than Apollo because it aggregates from local databases (Companies House, EU business registries). Use ONLY for UK T1 leads that failed both Prospeo and Apollo.
Go to kaspr.io. Log in.
Upload the UK-only Prospeo+Apollo misses CSV (filter the misses file to UK leads from Segments 1, 2, and 5 only).
Kaspr reveals emails from EU/UK databases.
Download results. Expected recovery: 10-15% additional UK emails.
~/Downloads/Phellos-Lead-Gen/4-Enrich/Phellos_Kaspr_UK.csvMerge all enrichment results back together and attach Crunchbase funding data.
Verify ALL emails before sending. Only "ok" status passes.
Go to millionverifier.com. Log in.
Click "Bulk Verifier" in the left sidebar.
Upload your enriched CSV (Founders or VC). MillionVerifier will ask which column contains the email addresses - select "email" (the column is named exactly email in all pipeline output files).
Click "Start Verification". This takes 5-30 minutes depending on list size.
When complete, download the results CSV.
Repeat for each enriched file.
~/Downloads/Phellos-Lead-Gen/5-Verify/T1 leads get Sherlock deep research and personalized hooks. T2 get rule-based hooks. T3 get generic templates. Scoring ensures you spend research time on the leads most likely to convert.
Founders Track Scoring Rubric (max ~185 points):
| Signal | Points |
|---|---|
| Founder + CEO combo | +25 |
| Founder / Co-Founder | +20 |
| CEO only | +15 |
| Signal | Points |
|---|---|
| SaaS / Software / Fintech / AI / Cloud / Data | +30 |
| Hardware-enabled tech | +30 |
| Adjacent (Marketing, Consulting, E-commerce) | +15 |
| Employees | Points | Note |
|---|---|---|
| 11-50 (Leo's sweet spot) | +25 | 10-30 is ideal within this range |
| 2-10 | +15 | Too early but might be interesting |
| 51-200 | +10 | Only if under $40M revenue |
| 200+ | 0 | Too large - "just hire FTEs" |
| Founded | Points | Note |
|---|---|---|
| 2022-2024 (2-3 years old) | +25 | Leo's stated sweet spot |
| 2020-2021 (4-5 years) | +15 | Acceptable |
| 2019 (6 years) | +5 | Only if actively mid-round |
| Before 2018 | 0 | Red flag per Leo |
| Location | Points |
|---|---|
| UK (Leo's primary market) | +15 |
| USA | +10 |
| Other English-speaking | +5 |
| Signal | Points |
|---|---|
| Series A or B raised in last 12 months | +40 |
| Series A or B raised 12-24 months ago | +35 |
| Seed in last 12 months | +20 |
| Seed 12-24 months ago | +10 |
| Pre-Seed in last 12 months | +15 |
| Pre-Seed 12-24 months ago | +8 |
| Total funding $1-50M | +10 |
| Signal | Points |
|---|---|
| Accelerator-backed (Seedcamp/Techstars/EF) | +15 |
| Signal | Points |
|---|---|
| Has company website | +5 |
| Has phone number | +5 |
| Tier | Score | Description |
|---|---|---|
| T1 | 120+ | Full ICP match - right funding, right age, right size, right industry |
| T2 | 80-119 | Good match - missing one dimension (e.g., right size but no funding data, or Pre-Seed only) |
| T3 | <80 | Volume play - generic outreach template |
Sales Nav leads without Crunchbase data cannot score funding (+40) or company age (+25) points. A perfect Sales Nav-only UK SaaS founder caps at ~100 points. To avoid systematically demoting these leads to T2, use adjusted thresholds:
| Track | T1 | T2 | T3 |
|---|---|---|---|
| With Crunchbase data (max ~190) | 120+ | 80-119 | <80 |
| Sales Nav only (max ~100) | 85+ | 55-84 | <55 |
last_funding_type, last_funding_date, AND founded_year are ALL empty, the lead has no Crunchbase data and the Sales Nav-only thresholds apply.VC Track Scoring Rubric (simpler, max ~100 points):
| Signal | Points |
|---|---|
| Fund invests in Leo's sweet spot (SaaS, Fintech, hardware) | +25 |
| Fund size/prestige signals (known firm) | +30 |
| UK or US based | +15 |
| Active on LinkedIn (Posted last 30 days) | +10 |
| Portfolio includes companies in Leo's ICP | +20 |
| Tier | Score | Description |
|---|---|---|
| T1 | 80+ | Strategic VC - invests in Leo's ICP sectors, active, known firm |
| T2 | 50-79 | Good VC - partial overlap with Leo's sectors |
| T3 | <50 | Low priority - generic outreach |
Jamie's build ran 95 Sherlock dossiers and used FireCrawl for website scraping. Both were overkill for podcast invites. This playbook caps deep research at 50 founders and skips FireCrawl entirely. Exa finds podcast appearances and press - that's all you need.
~/Downloads/Phellos-Lead-Gen/6-Sherlock-Dossiers/.Run Sherlock deep research on the top 50 T1 founders only. Not VCs - VC outreach uses a simpler referral template that doesn't need dossier-level personalization.
What Sherlock does: Takes each T1 founder's enriched data as starting context, then independently researches them using web search and Exa. Finds podcast appearances, press coverage, funding announcements, and pivot stories. Writes a full intelligence brief.
What it does NOT do (v2 change): No FireCrawl website scraping. No company hiring analysis. These are overkill for a podcast invite. Exa finds podcast appearances and press - that's the hook you need.
Cap: 50 leads maximum. ~3 minutes per lead = ~2.5 hours total.
~/Downloads/Phellos-Lead-Gen/6-Sherlock-Dossiers/. Runtime: ~2.5 hours (3 min/lead x 50 leads). Start this before lunch or before end of day - let it run in the background. It prints progress every 10 leads so you can check back.Research each T1 lead's biggest CLIENT or CUSTOMER - not their supplier, not their investor. This creates a powerful subject line variable that dramatically increases open rates.
For VCs, the "notable reference" is one of their portfolio companies - the one they're most publicly associated with.
subject_hook. Don't force a reference - a wrong client name is worse than no reference at all.biggest_client column. Expect 40-60% fill rate for founders (many SaaS companies list customers), higher for VCs (portfolios are usually public).Two hook sources: Claude Haiku generates hooks from Sherlock dossiers (top 50). Rule-based fallback covers remaining T1 leads without dossiers.
| Priority | Signal Detected | Hook Pattern | Notes |
|---|---|---|---|
| 1 | yc, techstars, seedcamp (accelerator_backed tag) | {firstName} post-accelerator | Most specific signal - always wins if present |
| 2 | raised, funding, series A (last_funding_type = Series A) | {firstName} post-series-a | Funding recency is Leo's top ICP signal |
| 3 | raised, funding, series B (last_funding_type = Series B) | {firstName} after the raise | |
| 4 | seed (last_funding_type = Seed) | {firstName} post-raise | |
| 5 | hardware, physical product (industry = hardware-enabled) | {firstName} hardware story | Distinctive vertical - higher than fintech/AI |
| 6 | fintech, financial (industry contains Financial Services) | {firstName} in fintech | |
| 7 | ai, machine learning (headline or title contains AI/ML) | {firstName} and ai | |
| 8 | scaling, growth (headline contains scaling/growth) | {firstName} scaling story | |
| 9 | uk, london, british (geography = UK with no stronger signal) | {firstName} uk tech | Last resort before default |
| Default | No signal match | quick question {firstName} | Used when no data signals are available |
| Rule | Requirement |
|---|---|
| Length | 20-30 words |
| Must start with | you, your, you've, you're, caught your, saw the news |
| No em dashes | Use regular dashes only |
| No non-ASCII | Strip smart quotes, accented characters, emojis |
| No garbage patterns | Reject: "been following", "love what you", "impressed by", hashtag spam, URL-only content |
T1: ~230 at 15-25% = 35-58 replies. T2: ~1,300 at 5-10% = 65-130 replies. T3: ~500 at 2-5% = 10-25 replies. VC: ~1,700 at 3-7% = 51-119 replies. Total: 160-330 conversations across both tracks.
[Mon Year] format (e.g., "Mar 2026").| Campaign Name | Tier | Est. Leads | Personalization |
|---|---|---|---|
| Phellos - SaaS UK T1 Pers - [Mon Year] | T1 | ~80 | Sherlock dossier + Haiku hooks (SaaS/hardware UK founders with dossiers) |
| Phellos - SaaS US T1 Pers - [Mon Year] | T1 | ~100 | Sherlock dossier + Haiku hooks (SaaS/hardware US founders with dossiers) |
| Phellos - Fintech T1 Pers - [Mon Year] | T1 | ~30 | Sherlock dossier + Haiku hooks (Fintech founders with dossiers, UK + US) |
| Phellos - SaaS T1 - [Mon Year] | T1 | ~320 | Rule-based hooks (SaaS/hardware UK + US founders WITHOUT dossiers) |
| Phellos - Fintech T1 - [Mon Year] | T1 | ~50 | Rule-based hooks (Fintech founders WITHOUT dossiers) |
| Phellos - SaaS UK T2 - [Mon Year] | T2 | ~400 | companyName only |
| Phellos - SaaS US T2 - [Mon Year] | T2 | ~600 | companyName only |
| Phellos - Fintech T2 - [Mon Year] | T2 | ~300 | companyName only |
| Phellos - Volume T3 - [Mon Year] | T3 | ~500 | companyName only |
| TOTAL (Founders Email) | ~2,380 | 9 campaigns |
| Campaign Name | Tier | Est. Leads | Personalization |
|---|---|---|---|
| Phellos - Venture Capital - [Mon Year] | T1/T2 | ~500 | companyName + fund focus |
| Phellos - Investment Management - [Mon Year] | T1/T2 | ~400 | companyName + fund focus |
| Phellos - Capital Markets - [Mon Year] | T1/T2 | ~500 | companyName + fund focus |
| Phellos - M&A Advisory - [Mon Year] | T1/T2 | ~300 | companyName + fund focus |
| TOTAL (VC Email) | ~1,700 | 4 campaigns |
| Column | Required | Notes |
|---|---|---|
email | Yes | Verified email address |
firstName | Yes | First name, cleaned of LinkedIn artifacts |
lastName | Yes | Last name |
companyName | Yes | Company name, truncated at 50 chars |
title | Yes | Job title |
linkedin_url | Yes | Full LinkedIn profile URL |
subject_hook | T1 only | Personalized subject line variable |
personalization_line | T1 Personalized only | Opening line from Sherlock/Haiku |
tier | Yes | T1, T2, or T3 |
website | Recommended | Company website |
company_linkedin | Recommended | Company LinkedIn page URL |
| Column | Required | Notes |
|---|---|---|
first_name | Yes | Note: underscore format, not camelCase |
last_name | Yes | |
linkedin_url | Yes | Full LinkedIn profile URL |
company_name | Yes | |
title | Yes | |
personalization_line | T1 only | |
tier | Yes |
Run this validation BEFORE uploading any CSV to Instantly or HeyReach.
| Field | Check |
|---|---|
| Name (first + last) | No blanks, no single characters |
| No blanks, valid format | |
| Title | No blanks |
| Company | No blanks |
| LinkedIn URL | No blanks, starts with linkedin.com/in/ |
| Tier | Must be T1, T2, or T3 |
| Field | Threshold |
|---|---|
| Website | Warn if <80% filled |
| Company LinkedIn | Warn if <80% filled |
| Check | What to Look For |
|---|---|
| Blank emails | Any row with no email address |
| Blank names | Any row with no first or last name |
| linkedin.com in website column | LinkedIn URL accidentally in website field |
| /in/ in company_linkedin column | Personal LinkedIn URL in company field |
| Em dashes in personalization | Non-ASCII characters trigger spam filters |
| Non-ASCII in any field | Smart quotes, accented chars, emojis |
| Invalid tier values | Anything other than T1, T2, T3 |
| Cross-file email duplicates | Same email in multiple campaign CSVs |
| Turkey company check | Final safety net - Turkish companies in any file |
~/Downloads/Phellos-Lead-Gen/
1-Raw-Scrapes/ # Raw HeyReach CSVs (one per sub-search)
2-Crunchbase-Sources/ # Raw Crunchbase CSVs (Seed, Series A, Series B)
3-Dedup-Clean/ # Deduped + title-filtered + DNC-filtered
4-Enrich/ # Prospeo + Apollo enrichment results
5-Verify/ # MillionVerifier verified + scored + tiered
6-Sherlock-Dossiers/ # Intelligence briefs for top 50 T1
7-Scripts/ # build_campaigns.py, generate_hooks.py
8-Campaign-CSVs/ # Final campaign CSVs
final/ # Upload-ready files for Instantly + HeyReach~/Downloads/Phellos-Lead-Gen/8-Campaign-CSVs/final/. Ready for upload.Instantly does NOT enforce a workspace-level daily sending limit. If campaign limits sum to more than 18 accounts x 30/day = 540 emails/day, Instantly WILL send more and destroy domain reputation.
Rules: (1) Sum of all campaign daily_limit values must equal 540. (2) Each account appears in only ONE campaign. (3) No account sends more than 30/day.
These alerts generate fresh T1 leads continuously between batch pipeline runs. Set up once, then check daily.
Go to google.com/alerts. Sign in with your Google account.
Create alert: UK SaaS startup Series A 2026 - set delivery to "Once a day", email.
Create alert: UK fintech startup funding 2026 - set delivery to "Once a day", email.
Create alert: London startup raised Series A - set delivery to "Once a day", email.
Create alert: UK startup seed funding 2026 - set delivery to "Once a day", email.
When alerts arrive: Check each company against Leo's ICP (2-3 year old, 10-50 employees, SaaS/Fintech/hardware, UK/US). If they qualify, add the founder directly to the Founders T1 campaign in Instantly.
These companies are Phellos existing clients. Any lead at these companies must be removed at Step 3 and re-checked at Step 8 quality gates.
| Company | Company | Company |
|---|---|---|
| ARF | Base64.ai | BasicBlock |
| Beije | Delivers AI | Denebunu |
| EaseGlobal | FineDineMenu | Folia Health |
| ForInvest | Guul Games | HelloKlean |
| Integrated Finance | Interzoon | Lidio |
| Mindsite | Mobile Commons | Shuma Labs |
| WOMO / WAMO | TrioMobil | Mindsight |
| Monay | Octopus | Oplog |
| Paket Mutfak | Rudiq | Teldio |