3 Minutes Guide: How To Scrape The YC Startup Directory Without Code

To scrape the YC startup directory without code, first generate a list of company-profile URLs from a filtered directory page, then process those URLs in a separate workflow to extract company and founder details. In Datablist, use AI Agent - Site Scraper for the listing stage and AI Agent - Run on Collection Items for the profile stage.

I recommend separating those stages because a directory card and a full company profile do not contain the same fields. Keep the YC Company URL through both phases so a founder or website stays attached to the right company.

This guide gives you two prompts, the output fields to configure, and checks for missing or repeated records. A template can supply a starting configuration, but it does not guarantee complete directory coverage or accurate founder details.

Start with the filtered directory, generate the profile URL list, then extract and review details.

Scraping The YC Startup Directory: The 2 Phase Process

Define one record before you configure the source. In Phase 1, one row represents one startup listing. In Phase 2, one row represents that company and the founder information available on its profile.

Finding The Most Relevant YC Startups

Open the Y Combinator company directory, apply the filters relevant to your project, and copy the resulting URL. Reopen that URL to confirm that it preserves the intended search and filters.

For example, define a batch and industry relevant to your product before collecting URLs. A directory label is a starting qualification signal, not proof that every company is active, growing, or ready to buy.

Keep the filter choices and collection time with the batch. Test whether the requested records are visible to the scraper. A dynamically loaded directory can show different content in a browser and in a fetched page; do not invent a next-page URL or assume scrolling means every record was loaded.

For a large inventory, use identifiable filtered batches and compare overlaps by YC Company URL. A row limit or a successful request is not evidence that the whole directory was collected. This guide does not rely on a fixed directory size or a universal results ceiling.

Phase 1: Scraping Startup Listings From The YC Directory

The first phase involves extracting the basic information from the YC directory search results. This gives you a foundation list of companies to work with.

Phase 1 - Step 1: Sign up & Create a Collection

First, sign up for Datablist.com

Then, create a New Collection

Phase 1 - Step 2: Create Your First Property to Prevent Duplicates

There are some startups in the Y Combinator directory that appear more than once in search results. If you don't want to have the same company multiple times in your collection, do the following:

1. Create a New Property and name it "YC Company URL"

2. Click on the Column Header and select Rename - Settings - Delete

3. Enable Do not allow duplicate values, then click on Save Property

Phase 1 - Step 3: Navigating To the AI Scraping Agent

1. Click on Import in the top menu of the app and select Import From Data Sources

2. Scroll down and select AI Agent - Site Scraper

Review the source settings before entering the listing URL. The screenshot illustrates the workflow; controls can change.

Phase 1 - Step 4: Selecting the Template

Start from a suitable directory template when available, or configure Url to scrape and Prompt manually using the example below. Review the selected settings rather than assuming a saved template fits the current page.

Paste the filtered directory URL into Url to scrape. Leave Enable Pagination off for the first sample.

Configure Expected Item Outputs to match the fields in the prompt. After the sample is useful, inspect the actual pagination pattern before enabling it and setting Max Pages. A missing Next link needs investigation; it is not proof that the directory has ended. Then choose Continue.

Use this prompt for the current listing page. It collects profile links before asking for details that may only exist on those profiles:

The directory URL is a separate setting. Map Startup Name and YC Company URL as text and URL outputs, plus Listing Description when useful. Retain Page Scraped to identify where a listing was found.

Phase 1 - Step 5: Selecting Inputs & Starting The Scrape

Check that YC Company URL maps to the intended URL property. A similar output name alone is not proof that the mapping is correct.

For the other properties, click on the ⊕ Icon to add them to your collection.

Once you've done that, click on Run Import Now to start scraping.

Inspect the generated rows against the directory before extending the run. The screenshot below shows the output layout rather than a promise of count or completion time.

You should now have:

• Startup Name
• YC Company URL
• Listing Description, when requested and available
• Page Scraped for provenance

You can stop here, or continue with the phase 2 to get information about the founders.

Phase 2: Scraping Company and Founder Details

In this phase, process the reviewed YC Company URLs from Phase 1. Choose fields the profile actually displays, such as Startup Name, Website URL, Batch, Location, Team Size, and Founder Details. Missing data should remain missing rather than being inferred from search snippets or a company description.

For a company-level record, Founder Details can retain each displayed founder's name and corresponding profile links together in one text field. If your downstream task needs one contact per row, configure that as a separate record shape and keep YC Company URL as the join key. A template with three founder slots is a chosen output layout, not a maximum number of founders a company can have.

Insert the YC Company URL property from the source collection at the end of the prompt. In Expected Item Outputs, configure the same names and types: URLs for links, text for names and labels, and long text for Founder Details. Keep Team Size as text when the page supplies a range or qualified value.

Phase 2 - Step 1: Creating a New Collection

Create a New Collection

Use a separate results collection so generated detail records do not replace the source URL list.

Phase 2 - Step 2: Selecting the Template

Select AI Agent - Run on Collection Items as the source for the new results collection.

Review the source configuration before entering the profile prompt.

Use an appropriate saved configuration when available, or paste the profile prompt above and configure its outputs. Inspect any template field names before running.

Phase 2 - Step 3: Mapping Your Collection & Input Property

In For every item in, select the reviewed listing collection from Phase 1.

At the end of the prompt, insert the YC Company URL property from that collection. Keep the literal # YC Company URL separator above it.

Type / and select YC Company URL to insert the property.

Once the input is mapped, click on Continue.

Phase 2 - Step 4: Selecting Outputs

Map every requested output to the intended property in the results collection. Include YC Company URL so details remain traceable to the source record.

For the data points you don't want, click on the ✕ Icons to remove them from your collection.

Once you've done that, scroll down and click on Run Import Now to start scraping.

Review a bounded source collection first, with examples of the profile formats in your project. The screenshot shows a result layout; it does not establish accuracy or completion time.

Compare returned fields with the actual company profile. Inspect request errors and missing fields separately. Check that each founder link belongs to the person named beside it before accepting the record.

Using the Results in Datablist

Plan Usage From a Sample

The listing and profile stages do different amounts of work. Check current source settings, available credits, and reported sample usage for each stage before extending the run. Do not assume that a fixed number of startups always costs the same amount, especially when profiles require extra page work or fail.

Keep Configuration and Review Together

A prompt describes what to extract, while output settings determine the fields you can use. Compare both after editing a template. If you request Website URL in the prompt but do not map it, the downstream list will not contain the expected field.

I would sample companies with missing websites, multiple founders, and differing profile layouts. Review both the first-stage profile links and the second-stage details before treating the batch as a lead list.

Enrich Accepted Contacts Separately

The directory scrape does not establish a founder's work email or current buying intent. After reviewing person identity and employer, use Waterfall Email Finder when you need a work-email lookup. Keep the original company URL and accepted founder information with that result.

Use waterfall enrichment to distinguish lookup results, verification information, and missing inputs. Deduplicate repeated company URLs across batches before creating repeated contact lookups.

Export a Traceable Startup List

Keep YC Company URL, source filters, collection time, reviewed company fields, and unresolved outcomes through the export. One repeated company found in two batches should not silently become two independent companies.

These examples describe the intended review workflow, not measured scrape results. Use CSV and Excel exports to hand off accepted records. If your goal is outreach, qualify the company and contact separately before using the list.

Frequently Asked Questions (FAQ) About Scraping The YC Startup Directory

How Much Does It Cost to Scrape The YC Directory?

Check the selected source and reported sample usage. Listing extraction, profile processing, model choices, and additional enrichment have different usage. This guide does not promise a fixed credit range for 500 startups.

How Can I Scrape Startups From The YC Directory?

Collect profile URLs from a filtered directory page with AI Agent - Site Scraper, then process accepted URLs with AI Agent - Run on Collection Items. Match prompts and outputs in each phase, review a sample, and export the useful records with their source URLs.

Does a Public Listing Guarantee Access or Complete Coverage?

No. A browser-visible directory does not establish that a scraper can access every record or that every listed field is current. Review the site's access requirements and terms for your intended use, check a sample, and keep the collection boundary explicit.

What Kind of Data Can I Get From The YC Directory?

Configure fields actually displayed on the listing or company profile, such as company name, website, batch, location, team size, and paired founder names and links. Output names do not guarantee that those values exist. Leave missing information empty.

Why Would I Want to Scrape YC Startups?

The directory can supply candidate companies for market research, recruiting, or a relevant sales project. Qualify them against your actual criteria. Membership in a startup directory does not prove current funding, growth, or readiness to buy.

Can I Customize What Data Gets Scraped?

Yes. Edit the prompt and expected-output settings together, then test the changed configuration. A field absent from the page needs a different research step; asking for it does not make it available.