> ## Documentation Index
> Fetch the complete documentation index at: https://docs.metal.so/llms.txt
> Use this file to discover all available pages before exploring further.
# AI Search
> Use historical data and investing patterns to surface the "most likely" investors for your specific Company.
Metal's investor search capabilities combine high-precision filters with AI-powered natural language search, creating two complementary ways to build a targeted list from 30,000+ active investors.
By observing large data sets across thousands of raises, we have learned that taking the time to empirically determine the "most likely" investors is the highest leverage activity in a raise process. \
\
Focusing on the strongest-fit investors improves intro request to opt in rates and 1st to call call conversion.
***
## Two Ways to Search
Build a custom search query using precision filters across stage, sector, geography, check size and more.
Use natural language to describe your ideal investor profile and let Metal's AI put in the right filters.
For most users, starting with natural language search is the right choice. For power users looking for customisation, setting the search criteria themselves can be a useful starting point.
***
## Precision Filters
### Investor Type
Filter by investor category: Accelerators, Angels, Corporate VCs, Family Offices, Private Equity, Venture Capital Firms, or Other. Use this to exclude investor types that structurally don't fit your raise — angels don't lead large rounds, and corporate VCs typically don't invest in early-stage companies raising pre-seed rounds.
### Investor HQs
Filter by the country or city where an investor is headquartered. Distinct from geography filters (which track where investors *deploy* capital) — HQs tells you where the firm is located.
### Stage Specialists
Filters investors based on the minimum percentage of total investments in a specific stage. If users set this high, then they will see firms that specialise within that particular stage.
Stage is the single most important filter. An investor who leads Series B rounds is not a relevant target for a pre-seed raise regardless of sector fit. Apply this filter first, before any others.
### Continent Focus
Filter investors based on the minimum number or percentage of their total investments in a specific continent. Use this to find investors with genuine geographic conviction — not just one opportunistic deal in a region.
### Country Focus
Filter investors based on the minimum number or percentage of their total investments in a specific country. More precise than continent-level filtering — useful when you need investors who actively operate in a specific market (e.g., Germany, UK, Brazil) rather than just occasionally backing companies there.
Use Continent Focus to build a broad geographic shortlist, then layer Country Focus to identify investors with a genuine presence in your specific market.
### Check Size
Filter investors based on the typical check size size (min–max). This can be a particularly useful fit signal (distinct from fund size).
### Fund Size
Filter investors by the size of their funds — a proxy for capacity and check size range. For most investors, the check size is about 1-2% of their fund size. Fund size can be used to identify specific types of investors. As an example, a \$50M fund is unlikely to write a \$5M cheque; a \$500M fund rarely leads a \$250K pre-seed.
Founders that are looking for follow-on investors into an existing round find the Fund Size filter to be particularly useful. It allows them to identify investors that have a small fund size and are therefore writing small follow-on checks (\$50-250K).
### Investment Activity
Filter investors based on 3mo, 6mo, 9mo or 12mo minimum deal count. High recent activity signals active deployment. Low recent activity may indicate a fund winding down, between vehicles, or highly selective. Founders typically prioritise high-activity investors when running a time-sensitive raise.
### Inclination to Lead
Filter investors based on the percentage of their total investments that they lead. When looking for a lead investor for the round, founders typically filter for investors with a meaningful historical lead rate of 20% or higher. A low lead rate doesn't disqualify an investor, but indicate that they're better positioned for a follow-on check.
### Inclination to Follow On
Filter investors based on how frequently they participate in follow-on rounds for companies that they previously backed. High follow-on rates signal commitment to portfolio companies — this is particularly relevant if you want investors who tend to support subsequent rounds, and not just write the first cheque.
### Sector Concentration
Filter investors based on the minimum *percentage* of total investments in a given parent or sub-sector. This is most useful in identifying investors that are concentrating investments in a specific sector.
| Threshold | What it surfaces |
| :------------------ | :----------------------------------------------- |
| 1–5% concentration | Generalists with passing sector familiarity |
| 5–10% concentration | Sector-aware investors with a developing thesis |
| 10%+ concentration | Sector specialists with deep category conviction |
### Sector Familiarity
Filter investors based on the minimum *count* of investments in a given sector. This is different from Sector Concentration: an investor with 2% concentration but 40 sector deals has more raw experience than one with 20% concentration and 3 deals. Most founders prefer to use the Sector Familiarity filter.
### Select Thesis
Multi-select filter from Metal's predefined thesis taxonomy. Covers thesis areas like Future of Work, Climate Tech, Developer Tools, FinTech Infrastructure, and others. Use this to find investors who have explicitly adopted a thesis aligned with your category — particularly useful in emerging verticals where sector tags alone may not capture the sub-market.
### Include Investors (by Tag)
Surfaces investors who have backed specific companies. The primary workflow is tag-based: build a tag list of comparable companies in Company Search, then apply that tag here to filter for investors who backed any company in your set. This is the highest-conviction signal in the filter set — covered in detail in [Similar Companies](/identifying-investors/similar-companies).
Avoid using category-defining companies (Stripe, Airbnb, Figma) as your comparable set. Their early investors operated at later stages and rarely match early-stage raises. Use companies that raised at a similar stage and round size to yours.
### Exclude Investors
Removes investors who backed companies you specify — most commonly your direct competitors. Investors with portfolio conflicts are unlikely to invest in you, and approaching them wastes time and can create awkward dynamics. Use this filter to clean your list before outreach.
### Co-Investors
Surfaces investors who have co-invested alongside a specific investor you select. Useful for mapping the syndicate ecosystem around a lead investor you're targeting — if you know Investor A is likely in your round, Co-Investors shows who regularly backs alongside them.
### Frequent Followers
Highlights investors who frequently participate in rounds where a specific lead investor has set terms. Useful for filling out a round once a lead is anchored — Frequent Followers tend to move fast when a trusted lead is already in.
***
## AI-Powered Natural Language Search
Type a description of your target investor into the global search bar — Metal's AI interprets the query and surfaces matching results by automatically applying the appropriate filters.
Example queries that work well:
> *"Seed investors backing B2B SaaS companies in Europe with a focus on vertical software"*
> *"VCs who have led rounds in AI infrastructure tools for enterprise"*
> *"Investors in climate tech who write checks between* $500K and $*2M"*
> *"Pre-seed funds that back technical founders building in developer tooling"*
The AI search draws on the same structured investor data as the filter search, but interprets intent from natural language rather than requiring you to pre-select filter values.
***
## Reading Results
Each investor card in search results shows the core signals you need to qualify quickly:
* **Name and Fund** — partner name and the fund they represent
* **Stage and Typical Cheque** — their standard investment parameters
* **Recent Activity** — deals closed in the last 12–24 months
* **Sector Tags** — their stated and revealed investment focus
* **Relevance Score** — Metal's ranking of how closely they match your company profile
* **Network Proximity** — how reachable they are through your existing connections
Click any investor card to open their full **Investor Profile** — portfolio history, thesis detail, valuation ranges, lead/follow behaviour, and intro pathways.
***
## Adding to Pipeline
From any search result, click **Add to Pipeline** to move an investor into your fundraising pipeline. You can set their initial stage — Identified, Researching, Outreach Pending — at the point of saving.
Don't filter too aggressively upfront. The goal of the discovery phase is to build a broad qualified list of 50–150 investors that you then qualify further through profile review and network mapping. Removing investors too early based on a single filter criterion means missing strong candidates.
***
Use AI Search alongside [Market Signals](https://docs.metal.so/Product-Docs-v2/Identifying%20Investors/market-signals), [New Funds](https://docs.metal.so/Product-Docs-v2/Identifying%20Investors/new-funds), and [Similar Companies](https://docs.metal.so/Product-Docs-v2/Identifying%20Investors/similar-companies) to build a multi-signal investor list that zooms in on the "most likely" investors.
***