Chapter 2
Sourcing Edge
Building Your Sourcing Edge
The “Bankable Thesis” Framework (Why you? Why now? Why this strategy?)
Mandate Alignment: Matching thesis to LP allocation priorities
Winning your first or second fund in 2025 isn’t about louder branding, it’s about a visible, repeatable edge. Founders are exhausted by spray-and-pray outreach and slow committees; LPs are screening for managers who can show unfair access to deals, clean underwriting, and institutional operations on day one. This guide gives you a practical playbook to build that moat: a relationship-led sourcing engine with operators, accelerators, angels, CVCs, universities, measured by conversion, an insight loop that publishes narrow theses founders actually use, a process that moves from intro to decision in 8-12 weeks, and an AI stack that compresses research, diligence, and portfolio monitoring. Wrap it in real controls, AML/KYC, cyber, valuation, conflicts, ILPA-style reporting, and you’ll do more than tell a story; you’ll run a machine LPs can underwrite.
The Sourcing Advantage
Founders are drowning in cold pings and “quick intros” that go nowhere. Decision cycles at large firms can stretch for months, forcing teams to run parallel processes and burn time. That fatigue creates a real opening for emerging managers who are narrow in scope, quick to decide, and genuinely helpful. If you can consistently deliver faster feedback, targeted customer intros, and crisp term sheets, founders will route you deals before the big brands even return an email.
LPs translate this into a single screening question: Do you have unfair access, and can you prove it? A credible “yes” never rests on vibes. It looks like a visible system:
- A repeatable sourcing engine
- Measurable relationships that actually produce warm intros
- Pattern recognition captured in your pass notes and investment memos.
In practice, that means your CRM and data room tell the same story the deck does, no hand-waving.
Start with repeatability. Map your channels such as operators, accelerators, angels, CVCs, and university programs, and tag every inbound with the true source. Show conversion by channel (intro → first meeting → diligence → term sheet → investment) and the median time from first touch to decision. Publish an internal SLA (e.g., 8-12 weeks to final answer), then measure yourself against it. Speed is a moat only if you can prove it.
Make relationships legible. List named operators who produce introductions each quarter; note the cadence and the average quality of referrals. With accelerators, specify batches covered, post-demo engagement, and what you offered founders with pricing workshop, buyer panel, regulatory templates. For CVCs and strategics, keep a one-pager on shared pipeline reviews and co-invest rules so you’re not negotiating from scratch each time.
Codify pattern recognition. A good deal log doesn’t just track passes, it explains why: channel economics, payback math, regulatory friction, founder/market mismatch. Over a few hundred reviews, those reasons become heatmaps that improve your hit rate and help you articulate what you’ll never back. That clarity is magnetic for founders and calming for LPs.
Finally, reduce friction for diligence. Organize your room to the headings allocators already use: a clean executive summary, a pipeline section with channel metrics and velocity, a track-record or case-study section with dated interventions and outcomes, and reporting samples formatted the way institutions expect. When your sourcing narrative is mirrored by hard numbers, artifacts, and a tidy room, you stop “selling” access, and start demonstrating it.
Sourcing Moat #1: The Network
Operator bench
Don’t just name “advisors”, treat operators like a quota-carrying channel. Build a roster of ex-CEOs/CTOs/VPs by domain (payer sales, industrial GTM, logistics, etc.) and assign each a light SLA: one founder roundtable per quarter, two reference calls per month, and one product/buyer intro per quarter. In your CRM, tag every intro with Operator_Name, Function, Company Archetype, and track conversion (intro → first mtg → diligence → term sheet → win). Share a simple quarterly leaderboard so the bench sees what works and where you need different expertise.
Operator bench
Don’t just name “advisors”, treat operators like a quota-carrying channel. Build a roster of ex-CEOs/CTOs/VPs by domain (payer sales, industrial GTM, logistics, etc.) and assign each a light SLA: one founder roundtable per quarter, two reference calls per month, and one product/buyer intro per quarter. In your CRM, tag every intro with Operator_Name, Function, Company Archetype, and track conversion (intro → first mtg → diligence → term sheet → win). Share a simple quarterly leaderboard so the bench sees what works and where you need different expertise.
Accelerators & programs
Publish a coverage map: which YC/Techstars/500 Global batches you track, who attends each demo day, and your post-batch offer with office hours, pricing workshop, buyer panel. For each program, log: Batch_ID, Companies contacted, Meetings set, Diligence started, Invested, and Avg time to decision. A one-page “Program Scorecard” in your data room proves this isn’t trophy-hunting, it’s a measured deal source. Bonus: run “theme days” and invite program teams + your operator bench to compress diligence.
Angel syndicates
Identify 15-30 high-volume angels in your vertical and put them on a monthly rhythm: a short “2 asks / 2 offers” email (what founders need; how you can help) and a tracked referral link that writes into your CRM with Angel_Source tags. Create a lightweight “angel pack” (how to qualify what’s on-thesis, sample questions, your kill-criteria) so referrals come pre-screened. Close the loop quickly, send a 3-line status within five days (“took a meeting / passed because X / moving to diligence”) to earn repeat flow.
CVCs & strategics
Treat corporate venture arms as dual channels: capital and customer access. Propose a quarterly pipeline sync where you share theme heatmaps and they surface internal problem statements. Agree a simple co-invest protocol with rights, info sharing, pilot NDAs, conflict boundaries and store it as a reusable one-pager. Track CVC-sourced deals separately, often the value is pilot velocity, so show days from intro → design partner → paid pilot → production.
Universities & venture labs
Pick the 3–5 labs that actually produce companies in your lane. Assign a partner “owner,” hold monthly office hours, and sponsor an annual prize tied to your thesis. In your dashboard, report Lab → Meetings → Diligence → Wins just like an accelerator.
Sourcing Moat #2: The Insight
You don’t win by guessing the future; you win by seeing the present more clearly and operationalizing that clarity. Insight should change who you meet, what you ask, and how you help, then show up as artifacts LPs can verify.
Market thesis
Write a one-pager per thesis with four fields: Claim, Counter-thesis, 5-day validation plan, Kill-criteria. Example: “Medicaid-adjacent digital health has a 10-year wedge in benefits navigation; we can validate payer willingness in five days via three buyer calls and two claims data partners.” The kill-criteria might read: “If two buyers reject for the same TAM/ops reason, pause the seam.” Date each revision so LPs see learning velocity. Add a buyer map and a distribution map with top channels, CAC/payback hypotheses.
Content-first signal
Ship a monthly memo that teaches one seam with receipts, screenshots, KPI ranges, procurement tactics, compliance checklists. Close each memo with “How we’ll test next month” and “What we’re not pursuing.” In the data room, mirror your external voice with ILPA-aligned reporting samples and a short Research SOP. This reduces diligence friction because LPs can connect your public insight to your internal operating system.
Question bank
Codify the 8-12 questions you can answer inside five days that most peers take weeks to crack with payer validation path, channel unit economics, regulatory gating items, data rights, switching costs. Tie each question to a toolkit with template buyer email, a 30-minute pricing survey, a data schema for cohort analysis). Keep timestamps: Day 1 outreach → Day 3 signals → Day 5 go/no-go.
Evidence of trusted-advisor status
Track engagement beyond the pitch: repeat founder consultations, diligence sessions where teams share KPIs, and explicit requests for customer intros or board/observer help. In your CRM, tag these as Advisor_Moments and summarize quarterly: “12 repeat consults; 7 KPI reviews; 5 pricing workshops; 4 buyer intros → 2 paid pilots.” In case studies, list dated interventions and the resulting KPI shifts.
Insight → sourcing loop
Every memo should generate a shortlist such as 10 founders, 5 buyers, 3 partners. Log outreach and track conversion from “read memo” → “took meeting” → “shared KPIs” → “term sheet.” When you can show that a thesis regularly yields warm intros and faster underwriting, you’ve turned opinion into a repeatable sourcing engine LPs can underwrite.
Sourcing Moat #3: The Process
Deal log
Run a single source of truth where every company reviewed is captured with date, source, theme, stage, ownership target, check size, IC owner, and a pass/advance reason coded from a tight list with distribution risk, CAC/payback, data rights, regulatory gating, founder–market fit, pricing power, unit-econ sensitivity. Add a short free-text note and link to artifacts such as buyer call notes, quick model, security review. Over 100-300 entries, those structured reasons become pattern heatmaps you can show in IC and to LPs.
Channel attribution
Tag origin with mutually exclusive, collectively exhaustive values: Operator, Accelerator (YC/Techstars/500), Angel, CVC/Strategic, University/Lab, Inbound Content, Portfolio Referral, Banker/Advisor. Your dashboard should report conversion by channel (intro → meeting → diligence → term sheet → invested) and median days between steps. Reallocate time toward channels with the best win-rate × speed and publish this reallocation in a quarterly “sourcing review” so the team learns with you.
Pipeline tiers with dated next actions
Keep the pipeline groomed weekly:
- Tier 1 (invest-ready): term-sheet window 0–30 days; list 2-3 blocking questions and your D1–D5 validation steps.
- Tier 2 (seed-ready in 1-2 milestones): capture the milestone plan, for example, paid pilot, gross margin target) and a check-in date.
- Tier 3 (early traction): store the hypothesis you’ll revisit such as new buyer segment, regulatory change).
Every card must carry a Next Step + Date + Owner to prevent drift and to show LPs you nurture future allocations rather than churn through leads.
Velocity with governance
Set SLAs by theme maturity: 8-12 weeks intro→decision for net-new themes; ≤4-6 weeks for pre-mapped seams. Track median and 80th-percentile cycle times plus time-in-stage (meeting→diligence, diligence→IC, IC→TS). Publish a monthly “stuck list” and clear them in a standing Pipeline Council meeting.
IC gates and memo discipline
Define standardized gates – Screen → Diligence → IC1 (go/no-go to negotiate) → IC2 with a two-page memo required at IC1 (thesis, risks, buyer map, unit econ) and a sources/uses + plan at IC2. Attach Week-1 kill-criteria (what ends work early) to every memo to keep speed without sloppiness.
Data hygiene & reporting
Automate dedupe, enforce required fields, and snapshot quarterly stats: companies reviewed YTD, pass reasons distribution, channel conversion, and velocity. Drop a redacted pipeline export and a one-page dashboard in your data room. That’s how process turns from a promise into a quantifiable moat.
AI Workflow Integration
Set the spine: policies, roles, guardrails
Before you wire tools together, write a 2-page AI SOP: what data can/can’t enter models, who can run which tools, and where outputs must be stored. Add a human-in-the-loop rule for any investment, performance, or LP-facing content, and log reviewer name/date. Create a lightweight prompt library and version it like code so your team stops reinventing queries.
Founder sourcing & research
Pair a research co-pilot with a reasoning model such as Claude to compress top-of-funnel. Use the browser assistant to summarize long articles, extract company facts, and generate seam-specific shortlists. Then have Claude draft a one-minute brief per company: what the product does, customer segment, pricing clues, and 3 qualification questions. Save every brief to the CRM with a source URL and timestamp. Because some AI browsers introduce security/automation risks, confine them to read-only sessions and disable form submission; route any downloads through a sandboxed machine.
Signal detection
Create a daily Signals View in your CRM that pulls from Crunchbase/LinkedIn: Actively Hiring, new VP Sales/Eng, 5+ open roles, executive churn, recent financing, or sudden job-post velocity. Assign a numeric score (0-10) and auto-create tasks when a company crosses the threshold (e.g., 7+). Add negative signals so your outreach stays targeted. For each hit, generate a personalized opener: 3 lines, one fact from their hiring pattern, one question that tests fit, one offer.
Diligence automation
Standardize a two-page IC memo skeleton. Feed call transcripts, product docs, and customer notes to the model to draft the first pass. Require humans to complete five sections: buyer map, unit-economics table such as CAC, payback, gross margin, sales cycle, competitive set, key risks, and Week-1 kill-criteria. Add a macro that builds sensitivity cases as price -10% / +10%, win rate ±5 pts, support cost ±15%) and stamps the workbook and memo with the approver’s initials. Keep the workpapers in your /204-2-style folder, if an LP or regulator asks “how did you compute that,” you can produce it instantly.
Portfolio monitoring & LP comms
Use a portfolio system such as Kushim, Edda to collect monthly KPIs. Point a model at those feeds to:
- Draft the portfolio snapshot: wins, risks, cash runway
- Flag outliers: cohorts where churn or gross margin deviates
Assemble talking points for founder reviews
Pipe the same inputs to your LP portal to pre-fill quarterly updates and attach charts. Add guardrails for performance language such as net vs. gross, benchmark caveats so you don’t trip marketing rules; final review is mandatory.
Founder enablement
Maintain three reusable playbooks, Pricing, GTM Hiring, Buyer Discovery, as checklists and micro-templates. Use AI to tailor each playbook to a company’s segment: generate a sample ICP, discovery script, 90-day comp plan, and a first-pass pricing grid with fences. Track dated interventions (“2025-03: 6 buyer interviews; 2025-04: pricing test; 2025-05: win rate +6 pts”) so case studies write themselves.
Data room governance
Choose a VDR with SOC 2 Type II and ISO 27001/27701. Let the model auto-tag uploads to your room index (00-07), generate consistent file names, and draft doc abstracts, so LPs can skim faster. Add an automatic PII scrub step on uploads (names, addresses, bank details) and block anything sensitive from leaving the VDR. Drop a one-pager in /04-Policies describing encryption, access logging, incident response, and vendor reviews, that alone reduces weeks of back-and-forth.
Risk management
Turn on usage logging: who prompted what, which files were accessed, what changed. Run a monthly AI red-team hour: test the prompts for hallucinations, leaking internal data, or unauthorized performance claims; update prompts and SOPs accordingly. Keep a model register, so you can answer “what produced this paragraph?” without scrambling.
Operational Excellence
AML/KYC onboarding
Spell out the exact path an LP follows from “interest” to “admitted.” Name your stack as e-subs portal, ID verification, sanctions/PEP screening, beneficial-owner look-through and publish service levels: document collection in 3-5 business days; screening within 24-48 hours; exception handling within 72 hours. Include an exceptions matrix such as trusts, foundations, offshore entities and who resolves them as admin vs. counsel. If you use a digitized platform, e.g., for real-time status and reminders, show screenshots and the owner responsible for chasing stale tasks. Close with a data retention note what is stored, for how long, who can access.
Cybersecurity
Condense your cyber program to two pages:
- Controls: SAML/SSO, MFA, role-based access, device encryption, least-privilege, quarterly access reviews
- Data protection: encryption at rest/in transit, DLP on email/VDR, restricted MNPI handling
- Operations: vendor due-diligence cadence, annual tabletop and penetration test, patch SLAs
- Evidence, VDR certifications, for example, SOC 2 Type II, ISO 27001/27701, audit reports available under NDA.
Add a one-page incident playbook like RACI, 72-hour notification protocol, regulator/LP comms template.
Valuation methodology
Publish a valuation policy with frequency, governing body, methods by instrument, data inputs, and a challenge process. Include one redacted valuation memo linking inputs to mark changes and a pricing exception log for events between quarters. Align section headings to ILPA DDQ so LPs can map quickly.
Conflicts & co-invest
Provide a decision tree for co-invest allocation, side-letter/MFN policy, and GP commitment sources/uses. List disclosure triggers such as cross-fund, board seats, related-party vendors and show the LPAC notification template you’ll use.
Reporting automation
Show a monthly operating dashboard with MRR growth, gross margin, cash runway, hiring, churn and quarterly financial packs aligned to ILPA’s 2025 Reporting & Performance Templates. Note delivery standards: T+30 draft, T+45 final, and the portal LPs will use. Include a sample capital account statement and K-1 timeline if applicable.
Team cadence
Document a weekly IC with a published agenda with pipeline, stuck items, mark changes, decision logs, and “memo-by-default” culture such as 2-pager required for IC1; update after material events). Assign one partner as CRM owner with monthly data-quality audits. ODD teams fund systems they can verify, make yours visible.
The Emerging Manager Checklist: Sourcing + Operations
Sourcing strategy
Publish a 3-5 page “Sourcing OS” that defines lead vs. referral vs. inbound, lists every channel you cover as operators, accelerators, angels, CVCs, universities, content, and shows a simple funnel: Intro → 1st mtg → Diligence → IC → Term Sheet → Invested.
Channel focus
Pick the three channels that actually produce wins and track quarterly conversion + velocity in your CRM. Report intros per month, conversion, median days between stages, and the win-rate × speed score that determines where you spend time next quarter.
Operator advisory board
List 5-10 operators and their time commitments: e.g., two diligence calls/month, one hiring referral/quarter, one pricing session/half-year. Keep a shared calendar and log outcomes.
Deal log
Maintain a monthly export showing 100+ companies YTD with source tags and coded pass reasons as distribution risk, CAC/payback, data rights, regulatory, founder-market fit. Add links to buyer-call notes and your quick model.
Policies
Upload one-pagers for AML/KYC, valuation, conflicts, cybersecurity, each paired with a redacted sample memo such as a valuation memo, a conflicts disclosure.
How to Fundraise Your First Fund in 2025
Defining Your Investment Thesis
LP Targeting: Family Offices to Fund-of-Funds
Building Your Sourcing Edge
Running Your Fundraising Process
Fund Structure: 506(b), AIFMD & ELTIF
First Close: Funnel & Metrics
US Marketing Rules & SEC Compliance
The Final Close Checklist
EU Fundraising Routes
LP Reporting & Communication
Building Your LP Data Room & DDQ
Secondaries & Continuation Funds