A good investor list isn't the one with the most names — it's the one where every name actually backs your stage, sector, check size, and geography, with evidence to prove it. This guide walks through a 6-step framework: (1) define your investor ICP, (2) source across databases and the open web, (3) review each candidate for fit and evidence — not just field matches, (4) rank and cut to a focused shortlist, (5) get partner-direct contacts, (6) track and de-dup. Steps 2 and 3 are the slow, manual part — and they're exactly what an AI people-search tool like Lessie can take on, describing your target in plain language and using AI Review to score fit against your criteria before you ever open a spreadsheet.
Why targeting beats volume
A directory export is only a starting point. Before outreach, every investor still needs to be checked for stage, thesis, check size, recent activity, and the right partner. That review process is what turns a raw list into a targeted investor list — and it's slow to do by hand. The framework below is how to do it well; the last section is where a tool can take over the slow middle steps.
The 6-step framework
Step 1 — Define your investor ICP
Before you look at a single name, write down what "a fit" actually means for you. An investor ICP has five dimensions:
- Stage — pre-seed, seed, Series A? A fund's sweet spot matters more than whether they can do your round.
- Sector / thesis — not just "SaaS" but the specific thesis (e.g. AI infrastructure, vertical fintech, dev tools). A fund's stated thesis is the clearest signal of whether it would back what you're building.
- Check size — the range they typically write. A $250K–$1M writer and a $10M lead are different jobs.
- Geography — where they invest, which isn't always where they're based. Some funds only do their home region; some go global.
- Activity / recency — are they actively deploying now, or between funds? "Active in the last 12–18 months" is a real filter, not a nice-to-have.
Write this as one plain sentence you could hand to another person. If you can't say it in a sentence, your list will be vague — and a vague list is why outreach feels useless.
Turn those five dimensions into one reusable brief:
Find [investor type] investing at the [stage] stage in [sector/thesis], with typical checks of [check size], investing in [geography], and evidence of activity within [time period].
Example: "Find seed investors backing AI infrastructure startups in the US, writing $250K–$1M checks, with at least one relevant investment in the past 12 months."
This same brief can be entered directly into Lessie as a plain-language investor search.
Step 2 — Source across databases and the open web
Now find candidates. No single source is complete, so use both:
- Structured databases (Crunchbase, PitchBook, investor directories) recall funds, partners, and past deals fast and in a clean schema.
- The open web — fund websites, partner posts, podcast appearances, portfolio pages, recent news — surfaces investors who aren't well-covered in any database, and it's where you find the evidence (a stated thesis, a recent deal in your space) that a database row doesn't carry.
Doing this by hand means running the same Boolean query across several tools, then Googling each name to confirm they still invest and still care about your space. It's the most time-consuming step and the one people cut corners on.
Step 3 — Review each candidate for fit and evidence, not just field matches
This is the step that separates a targeted list from a big one. A database row can match every field — stage "seed", sector "AI", geography "US" — and still be a bad fit, because the fund's actual thesis moved on, their last AI check was three years ago, or the partner who did those deals left.
Reviewing for fit means, for each candidate, answering: does the evidence support that this investor would back a company like mine, now? That's reading the thesis, checking recent deals, confirming the stage and check size against real activity — and flagging the ones you're unsure about instead of pretending a field match is a fit. Done properly, this is slow. Done sloppily, an untargeted export still requires every candidate to be re-checked before outreach.
Step 4 — Rank and cut to a focused shortlist
A targeted list is short. Once you've reviewed candidates, rank them by fit (how well thesis, stage, check size, and geography line up) and recency (how recently they've deployed into your space), then cut hard. Build a focused, well-matched list rather than a large untargeted export — it's a list you can actually personalize and follow up on, which is the whole point.
Group the survivors into tiers (dream / strong / stretch) so you can sequence outreach and spend your best personalization on the names most likely to convert.
Step 5 — Get partner-direct contacts
A generic [email protected] inbox rarely reaches the partner who makes the decision. You want the specific partner who invests in your stage and sector, and a direct, verified email. Match the name to the person whose thesis you referenced in Step 3 — the same partner you'd want to reach is usually the one whose recent deal made them a fit. Always verify an address before a large send; a bounce hurts your sender reputation and you only get one first impression.
Step 6 — Track and de-dup
A raise runs for months and usually across more than one tool. Keep one source of truth for who's on the list, who's been contacted, when, and what they said. The two failure modes here are (a) contacting the same investor twice under different tools, which looks sloppy, and (b) losing track of who's due for a follow-up. De-dup across your whole pipeline, not just within one list.
Where an AI people-search tool does Steps 2–3 for you
Steps 2 and 3 — sourcing across databases and the open web, then reviewing every candidate for real fit and evidence — are the slow, manual heart of building a targeted list. This is exactly what an AI people-search tool automates, and it's what Lessie is built for.
Instead of learning filter syntax and running the same query across several tools, you describe your investor ICP from Step 1 in plain language — "seed investors backing AI infrastructure in the US, check size $250K–$1M" — and Lessie searches across structured databases and the open web in one pass (Step 2). Its AI Review step then scores each candidate against your criteria and shows the supporting evidence — thesis, recent deals, stage — so what comes back is a short list of genuinely relevant investors with the reasoning attached, not thousands of rows to sort by hand (Step 3). From there it carries into the rest of the framework: you can unlock partner-direct contacts and send personalized outreach in the same workflow (Step 5), and it can keep the goal running on a schedule — re-searching, following up, and helping prevent duplicate outreach — with an automatic or human-in-the-loop approval step before anything sends (Step 6).
Lessie ranked #1 overall in PeopleSearchBench, an external open-source benchmark covering 119 general people-search queries; the benchmark did not specifically evaluate investor search. Paid plans start at $99/month and it's free to start — see current plans at lessie.ai/pricing.
The honest framing: a tool won't define your ICP for you (Step 1 is still your judgment) or decide your final tiers (Step 4 is a call you own). What it does is collapse the sourcing-and-reviewing grind into something you review rather than perform.
Manual research vs Lessie: building the list
| Step | Doing it manually | With Lessie |
|---|---|---|
| 1. Define ICP | You write it down — a plain-language brief | Same — you still own this; the brief becomes the search input |
| 2. Source | Run Boolean queries across several tools, copy rows into a sheet, Google each name | Describe the ICP once; it searches structured databases + the open web in one pass |
| 3. Review for fit | Read each fund's thesis and recent deals by hand; guess on the unsure ones | AI Review scores fit against your criteria and attaches the evidence |
| 4. Rank & cut | Sort the sheet by hand; cut to a shortlist | Candidates come pre-scored for fit; you make the final tiering call |
| 5. Get contacts | Hunt for partner emails in a separate finder; verify each | Unlock partner-direct contacts in the same workflow |
| 6. Track & de-dup | Maintain a CRM by hand; risk double-contacting across tools | Candidate pool tracks status and helps prevent duplicate outreach across runs |
| Where your time goes | Sourcing and reviewing (Steps 2–3) are the bulk of the manual work | You spend your time on Steps 1 and 4 — the judgment calls |
The takeaway isn't "tool good, manual bad." It's that the parts worth your time are defining the target (Step 1) and deciding the final list (Step 4). The sourcing and reviewing in between is mechanical, repetitive work — the right place to let a tool carry the load.
常见问题解答
How do I build a targeted investor list by stage and sector?
How many investors should be on my list?
What makes an investor a "fit" beyond matching filters?
Where do I find investor contact information?
Should I just use a database instead of a tool that reviews fit?
Can I keep the list updated as I raise?
Sources
Last updated: August 2026. Lessie capabilities and pricing: lessie.ai/pricing and lessie.ai/investor-scouting. Benchmark: PeopleSearchBench, arXiv:2603.27476 (general people-search, not investor-specific).