How to find investors for your startup (2026)

Most founders treat finding investors as a database problem. It is a pipeline problem. The 2026 workflow: pick channels by stage, filter for the funds writing your check size, tier into A/B/C lists, then run a partner-level research pass before sending. That last step is where weak outreach gets sorted from real meetings.

Jump to:

• How to find investors for your startup in 7 steps
• Where to find investors for startup rounds: the 2026 channel stack
• How to find VCs by stage and check size
• How to find angel investors, scouts, and rolling funds
• Build your investor list with an A/B/C tier system
• How to research investors at the partner level, not just the firm
• How many investors to contact and what reply rate to expect
• The 2026 filter that comes before everything else
• FAQ

The question of how to find investors for your startup has a stale default answer: scrape a database, export a CSV, send 200 cold emails. In 2026 that workflow gets you a sub-2% reply rate and a four-month dead pipeline.

The real work sits upstream of outreach. Capital is bigger than it has ever been outside the 2021 bubble: US venture deal value hit $267.2B in Q1 2026 alone, the highest quarterly total in four years, propelled by mega-rounds including OpenAI's $122B financing (PitchBook-NVCA Q1 2026 Venture Monitor). That capital concentrates: 41.5% of AngelList deals in H1 2025 went to AI/ML, and AI plus robotics accounted for roughly 60% of all capital deployed on the platform (AngelList State of US Early-Stage Venture H1'25).

Translation: the money is there, but only for founders who target the right funds, the right partners, and the right channels. This guide is the workflow, in order.

How to find investors for your startup in 7 steps

The whole workflow, in the order you should run it:

1. Define your round shape. Round size, lead check, follower checks, runway, dilution ceiling. Without these numbers your list will be noise.
2. Pick the right channels for your stage. Pre-seed and seed pull from different sources than Series A. The channel stack is below.
3. Source a raw long-list of 150 to 300 funds and angels. Combine database outputs (OpenVC, Signal, Crunchbase) with portfolio mining (their LPs, their syndicates, their scouts).
4. Filter the long-list down to 60 to 120 real targets by stage, check size, sector thesis, geography, and recent activity in the last 90 days.
5. Tier the filtered list into A/B/C. Tier-A leads (15 to 25), Tier-B followers and second-choice leads (30 to 50), Tier-C long shots (20 to 40).
6. Run a partner-level research pass on every Tier-A name. The partner, not just the firm. This is the step everyone skips.
7. Send in waves. Tier-B first (price the round, learn the objections), then Tier-A (with leverage and a refined narrative), then Tier-C as backup.

The rest of this guide expands each step with concrete tooling, 2026 benchmarks, and the rules that have changed in the last 12 months.

Where to find investors for startup rounds: the 2026 channel stack

There are five channels worth knowing, and they are not interchangeable.

• VC database aggregators. OpenVC publishes a free directory of 53 VC databases sorted by region, stage, and sector. Pitchbook and Crunchbase Pro have richer coverage, but the paid tools start at four figures a year and expose the same data your competitors are mining.
• Portfolio mining. Pick five companies one stage behind yours that look like adjacent or competitive bets. Their full investor list is usually on the company's About page, on Crunchbase, or in a Signal profile. Every one of those investors is a warm proof point: they wrote a check into your space recently.
• Signal by NFX. The single best free tool for finding partners (not firms) by sector and stage. It maps which individual partners have led which deals, not just which firms participated.
• LinkedIn and X searches. Search the word "investor" plus your sector plus recent posts. Partners advertise their thesis in public. A partner who posted three times this month about agentic AI infra is a hotter target than a partner whose feed went silent in 2023.
• Syndicates and rolling funds. AngelList Rolling Funds, scout funds, and SPVs have absorbed a serious share of the seed checks that used to flow only through institutional micro-VCs. They are also faster (decisions in days, not weeks) and write smaller checks that round out a pipeline.

The order of operations matters. Mine portfolios first so your database queries are anchored in real comparable rounds. Then use OpenVC and Signal to widen the list. Then layer in syndicates as you go. The reverse order gives you a 300-name spreadsheet with nothing tying any name to your actual round.

How to find VCs by stage and check size

How to find VCs that fit your round is a check-size problem before it is a sector problem. A $5M fund cannot lead your $15M Series A. A $500M fund will not bother with your $400K pre-seed party round. Get the arithmetic right first.

The math is simple: a VC writes at most 2 to 3% of fund size into any single deal, leads at typical ownership of 10 to 20%, and reserves 1 to 2x their initial check for follow-on. Reverse it, and your round size plus the lead's ownership target tells you exactly which fund sizes can play.

The 2026 wrinkle: median ARR to raise a Series A reached $2.5M in 2024, 75% higher than 2021, and median VC-backed cash runway dropped to 12 months, the lowest since 2019 (SVB State of the Markets H1 2025). If your numbers are below the funnel into the next tier, you are wasting cycles on funds that will pass on traction, not thesis.

Filter your raw list against three hard cuts:

• Stage match: they have led at least 3 deals at your exact stage in the last 18 months.
• Check size match: their median first check matches the lead check you need (or you file them as Tier-B follower, not lead).
• Recent activity: they closed at least one deal in the last 90 days. A fund that has gone quiet for 6 months is either out of dry powder or fundraising itself; either way, they will not move on your round.

How to find angel investors, scouts, and rolling funds

To find angel investors at scale, mine the cap tables of companies one stage ahead of yours. Angels who put $25K to $100K into a company adjacent to yours 12 to 24 months ago are vastly more likely to write a check than a cold name pulled from a list of "top startup investors."

Three sourcing channels for angels:

• Portfolio mining (again). Same five comparable companies. Their angel list is usually thinner than their VC list and almost always missing from databases. SEC Form D filings (free, public on EDGAR) name every investor on a priced round. Even on SAFEs, the angels usually appear in press releases or LinkedIn announcements.
• AngelList syndicates and rolling funds. Browse by sector. A syndicate lead with a thesis match brings 10 to 50 LPs along, and you handle one relationship instead of fifty.
• Operator angels in public. Founders who exited in your space in the last 24 months are the highest-conversion angels per email sent. They understand the problem, write fast, and bring distribution. LinkedIn searches for "exited" plus your sector return more usable names than any paid database.

What to skip: AngelList's static "find investors" lobby pages, generic angel networks that charge to introduce you (the model breaks because the angel is paying for deal flow, which means dabblers, not check-writers), and "Top 100 angels of 2024" listicles (the names are right, but they are also being emailed by every other founder reading the same listicle).

Scout funds are a separate category and usually overlooked. Sequoia Scouts, Lightspeed Scouts, a16z Scouts: each scout has $50K to $250K of fund money to deploy on the firm's behalf, often into deals the firm itself would consider too early. Getting on a scout's radar is the cheapest known path into a top-tier firm without going through the front door.

Build your investor list with an A/B/C tier system

Build your investor list as three tiers, not one ranked column. Treating all 80 names as a single ranked pipeline is the most common mistake in founder fundraising hygiene, and it produces the worst outcome: you send the best version of your pitch to the names you reach first, not the names that matter most.

The tier system:

The rules that make this work:

• Tier-A is not the first emails you send. It is the most important pitch you make, and you only get one shot per partnership. Use Tier-B to refine the deck and the data room before any Tier-A meeting.
• No name lives in two tiers. If you cannot decide, it is Tier-B. Tier-A discipline keeps the meaningful pitches focused.
• Re-tier weekly. As you learn what is and is not resonating, names move. A Tier-B that signals strong interest moves to Tier-A; a Tier-A that ghosts after three nudges falls to Tier-C or drops.

YC's Geoff Ralston captures the logic: meet as many investors as possible but focus on those most likely to close. The tier system is the operational version of that line.

How to research investors at the partner level, not just the firm

The most underrated step in how to research investors is researching the partner, not the firm. Firms don't write checks; partners do. A partner's investment history, their public thesis posts, and the three to five deals they personally championed in the last 24 months tell you whether a meeting with them ends in a yes or a polite pass.

The 2026 partner-research pass takes 12 to 20 minutes per name and looks like this:

1. Partner name + firm.
2. The last 5 deals they personally led or co-led (Crunchbase,
   Signal, firm's portfolio page).
3. Sector breakdown of their personal portfolio (not the firm's).
4. Their public thesis: last 10 posts on X / LinkedIn / Substack,
   any podcast appearance in the last 6 months.
5. Mutual connections (LinkedIn, plus any portfolio CEO you know).
6. One specific reason this partner cares about your wedge: a
   quote from a post, a portfolio company adjacency, a public
   thesis line.

That last line is the entire output. If you cannot write one sentence on why this partner cares about this wedge, you do not have a Tier-A target, you have a Tier-C long shot.

First Round partner Liz Wessel goes further: she recommends every founder identify a point partner at each target firm and book a 30-minute prep call the day before the formal partner meeting to learn what context the other partners already have, which slides to emphasize, and where the partnership's hesitations sit (First Round Review). That prep call is only unlockable by partner-level research, not firm-level.

AI is what makes this scalable. In 2024, doing 20 minutes of partner research on 60 names was a full week of work. In 2026, well-built prompts on frontier models reduce it to 30 to 60 seconds per partner once you feed them the partner's recent posts, deal list, and your wedge. If you are sending more than 20 of these a week, tools like Causo automate the partner research pass and draft the outreach off it.

The founders who close rounds in 2026 are not the ones with the longest investor lists. They are the ones whose first email to a partner names a deal that partner personally led and explains, in one sentence, why their thesis is your wedge.

How many investors to contact and what reply rate to expect

Contact 60 to 120 investors total to close a seed round, not 200 and not 30. Below 60, you do not have enough top-of-funnel for one yes to unlock the rest. Above 120, you are sending undifferentiated outreach and your reply rate craters.

The funnel math, drawn from operator data and YC's published guidance:

To get one lead and 4 to 8 followers (a typical seed round), you need roughly 5 to 10 partner meetings, which means 25 to 40 first meetings, which means 80 to 120 contacted investors at typical cold conversion. If you have warm intros for half your list, the contact count drops toward 60.

Bias your contact mix toward warm where you can. Cold still works. OpenVC's 2026 cold email guide mandates a subject line under 60 characters, a body under 1,000 characters, and a confident close like "Looking forward to showing you how X will change Y." Vague closes ("let me know if interested") and generic openers ("Hi there,") tank reply rates the most.

YC's Aaron Harris reframes the whole funnel: the only thing that matters initially is getting a yes from one investor. The first close unlocks the rest, which is why Wave 1 plus Wave 2 sequencing matters more than total contact count.

The 2026 filter that comes before everything else

AI thesis fit is now the first filter, not the last. In 2024, CB Insights tracked AI startups capturing 37% of all global venture funding and 17% of all deals, the highest funding share on record, with 74% of AI deals at early stage (CB Insights State of Venture 2024). 2025 widened the gap: Atomico reports AI and deep tech captured 36% of European VC funding (State of European Tech 2025), and AngelList shows 41.5% of platform deals going to AI/ML in H1 2025 alone.

What this means operationally:

• If you are building in AI, the long-list of funds writing your check is roughly twice the size it was 24 months ago. Use that breadth, but also expect deal sizes to be inflated: median venture deal size hit a record $3.5M in Q1 2025 (CB Insights Q1 2025) and median seed post-money valuations on Carta reached $24M in Q4 2025 (Carta State of Private Markets Q4 2025). You can ask for more; you also have more competition.
• If you are not building in AI, recognize the funding compression. The remaining 60 to 65% of capital is splitting across fintech, healthtech, climate, vertical SaaS, and bio. Your list will be narrower and your sequencing more important. Lead with traction, not vision.

The other shift: 2025 saw $425B in global startup funding across 24,000+ companies, a 30% YoY jump and the third-biggest venture year on record (Crunchbase 2025 Year-in-Review). The market is not closed. It is just sorted.

Run your list through one final cut before any outreach: of your 80 filtered names, which 10 to 15 have made a thesis-fit investment in the last 60 days? Those are your warmest cold targets. Everyone else is a step colder.

FAQ

How do you find the right investors for your startup?

Start with your round shape (size, lead check, runway), then filter for funds whose stage, check size, and sector thesis all match. The right investor is one who has led at least three deals at your exact stage in the last 18 months and has a partner whose public thesis aligns with your wedge. Most founders skip the partner-level filter and end up with a list of firms instead of a list of decision-makers.

Where do you find angel investors?

Mine the cap tables of five companies one stage ahead of yours. Their angels appear in SEC Form D filings on EDGAR (free), in Crunchbase, on the company's About page, and in launch press releases. AngelList syndicates, scout funds, and operator-angels who exited in your sector in the last 24 months are the next three highest-yield channels. Generic "top angels" lists and pay-to-introduce networks are the lowest-yield channel by a wide margin.

How do you build a target investor list?

Source a raw long-list of 150 to 300 names from databases (OpenVC, Signal, Crunchbase) and portfolio mining, then filter to 60 to 120 real targets using three hard cuts: stage match, check size match, and activity in the last 90 days. Tier the filtered list into A/B/C: 15 to 25 lead targets, 30 to 50 followers or alt-leads, 20 to 40 long shots. Contact in waves, Tier-B first to price the round, then Tier-A with leverage.

How many investors should you reach out to?

60 to 120 for a typical seed round. The math works back from outcomes: you need 1 lead plus 4 to 8 followers, which requires 5 to 10 partner meetings, 25 to 40 first meetings, and therefore 80 to 120 contacts at cold conversion rates (or closer to 60 if half your intros are warm). Below 60 you lack top-of-funnel; above 120 you are sending undifferentiated outreach and your reply rate collapses.

How do you research a VC before pitching?

Spend 12 to 20 minutes per partner (not per firm) on the five things that predict conversion: the last 5 deals they personally led, the sector breakdown of their personal portfolio, their public thesis from the last 10 posts and any recent podcast, mutual connections, and one specific reason this partner cares about your wedge. If you cannot write one sentence on the last point, the partner is not a Tier-A target. Frontier-model AI tooling now compresses the per-partner pass from 20 minutes to 30 to 60 seconds when you batch the inputs.

Related on the hub

• Investor update template (monthly, seed stage) — Copy-paste monthly investor update template for seed founders, with TL;DR, KPI table, asks block, lo…
• How to cold email VCs in 2026: the tactical playbook — Related cold outreach guide.
• The VC fundraising process in 2026: inside the firm — Related vc process guide.
• The VC due diligence checklist for a seed round in 2026 — Related vc process guide.