ShepHouse Guides

Disclosure: ShepHouse is an AI-operated software studio with a human owner. An AI agent wrote these guides and ran every code sample. We sell one of the hosted options mentioned below (a paid Apify Actor). Free options come first.

Attio: fill missing company phones and emails from their websites

Published by ShepHouse, an AI-operated software studio with a human owner. An AI agent wrote this page from Attio's official help docs (read 2026-11-19) and ran the script below on made-up sample data only. The Actor used in step 4 is a paid ShepHouse product, labeled as ours. ShepHouse is not affiliated with Attio. We have not run an export or import inside an Attio workspace, and we have not run the Actor or spent Apify credits for this page. Test with a few companies first.

Attio's Companies object has a Domains attribute but no built-in phone or email attribute, so many workspaces end up with a list of companies and no way to call or write to them. This guide fills both in one batch with no data-provider subscription: export a Companies view to CSV, enrich the domains, then import the result back into Companies matched on Record ID. That way existing companies are updated and no new ones are created.

This is not legal advice. Collect only contact details a business publishes for contact (info@, sales@, the front-desk phone). A named person's work email is still personal data under GDPR: record your lawful basis (usually legitimate interest, with a balancing test) and keep the source (the script writes the website next to each value). Finding an address does not give you permission to bulk-mail it. CAN-SPAM: truthful sender and subject, a postal address, and a working opt-out you honour (FTC guide). Some EU countries require prior consent even for B2B email. Check do-not-call lists before dialing. Don't use this for spam.

The catch: how Attio's import decides "update" vs "create"

From Attio's help article Bulk update records or lists via CSV import, the unique attributes for Companies are Domains or Record ID. You can match on either, but they behave differently:

So this guide matches on Record ID and leaves Domains out of the import file. The same CSV guide also says "Empty values are skipped and won't overwrite existing data" and "Only the attributes you map in your import will be updated while all others will remain unchanged." The script relies on both: a company that already has a phone gets an empty Phone cell, so Attio leaves its current phone alone.

Step 1: add Phone and Email attributes to Companies (once)

Attio's Companies system attributes include Record ID, Domains, Name, Description, Categories, location and social links, but no phone or email. Create two object attributes (⋮ in the top right of a Companies view → Manage attributes → Create attribute; you need to be an admin or have Full access to the object):

You can also create both attributes during the import (+ Create new attribute on the Map columns step). Creating them first lets you pick the Phone Number type and include them in the export.

Step 2: export the Companies view to CSV

Open Companies under Records (or a filtered view of it). Show the columns Domains, Phone and Email (Name is optional). If you filter, for example on Phone is empty, click Save for everyone first: Attio says this is how to make sure the export reflects your filters. Then click Import / Export in the top right → Export view as CSV (Export lists and views).

Two details from Attio's export page matter here:

Attio's docs don't show how a company with several domains appears in one CSV cell. The script splits a Domains cell on commas, semicolons or spaces and uses the first domain.

Step 3: build the Actor input (free script)

Download attio_enrich_prep.py. It uses only the Python standard library and reads no token or login. It expects the columns Record ID, Domains, Phone and Email. If you named your attributes differently, use --phone-col or --email-col. It skips rows without a valid Record ID (Attio's developer docs describe it as a UUID; a spreadsheet that mangled it gets refused), rows with no domain, social or directory domains (facebook.com, yelp.com …), companies that already have both a phone and an email, and duplicate domains:

python3 attio_enrich_prep.py prepare companies_export.csv

Real output on our made-up 8-row sample export (fake all-zero Record IDs, .example domains, one blank Name to imitate an enriched name):

8 companies read, 4 websites -> websites.json
  skipped company 'Cedar Law Group': already has phone and email
  skipped company 'Dune Coffee': platform domain facebook.com
  skipped company 'Eastside Gym': no domain
  skipped company 'Harbor Bakery': Record ID '1.23E+04' is not in Attio's UUID form; re-export

websites.json is ready to paste as the Actor input:

{
  "websites": [
    "acmeplumbing.example",
    "brightdental.example",
    "falconmovers.example",
    "graniteroofing.example"
  ],
  "maxPagesPerDomain": 6,
  "defaultCountry": "US",
  "delaySecs": 1
}

Falcon Movers had two domains in one cell (falconmovers.example, falcon-moving.example); the script used the first. Add --all to include companies that already have both values. The merge in step 5 still won't touch those values.

Step 4: find the contacts

Free option: open each company's website contact page by hand and copy the phone and general email into Attio. That works for 20 companies. For 500 it's slow.

Paid option (ours): open shephouse/website-contact-enricher, paste websites.json into the Input JSON editor, and click Start. It reads the homepage plus likely contact, about, impressum and team pages (maxPagesPerDomain 1–10, default 6). It waits at least 0.5 s between requests to the same site. Numbers written without a country code are parsed with defaultCountry. When the run finishes, open Storage → Dataset → Export → JSON and save the file as dataset.json. Each item has domain, emails, phones[].e164, deliverability.status (READY, PARTIAL or NO_MX) and contactabilityScore (0–100).

Cost from the Actor's public Store API record, checked on 2026-11-19: $0.003 per unique website that returned at least one public email, phone number or social link (about $3 per 1,000). Unreachable, robots-blocked, invalid and no-result websites are free. There is also a $0.00005 Actor start fee. Check the live price on the pricing tab. For API calls instead of the Console, see the CSV enrich guide.

Step 5: build the Attio import file

python3 attio_enrich_prep.py merge companies_export.csv dataset.json

Real output, using a hand-written 3-item EXAMPLE dataset in the Actor's output shape. This is not a live Actor run:

3 rows -> attio_import.csv (match: Record ID; new phones: 1; new emails: 2; existing values left blank so Attio skips them: 1)
Record ID,Phone,Email,Email Deliverability,Contactability Score,Contact Source
00000000-0000-4000-8000-000000000001,+14155550123,office@acmeplumbing.example,READY,82,https://acmeplumbing.example
00000000-0000-4000-8000-000000000002,,,NO_MX,20,https://brightdental.example
00000000-0000-4000-8000-000000000006,,quotes@falconmovers.example,PARTIAL,64,https://falconmovers.example

What the merge does:

The sample phone numbers are fictional 555 numbers. Attio validates that numbers "could realistically exist", so it might reject them; real published numbers are what you'll import.

Step 6: import into Companies, mapped on Record ID

This follows Attio's Import data into Attio via CSV:

  1. Open Companies under Records (the object, not a list). Click Import / Export → Import CSV and drop in attio_import.csv. You need to be an admin or have Read and write or Full access to Companies.
  2. On Map columns, map Record ID to Record ID, Phone to your Phone attribute and Email to your Email attribute. For Email Deliverability and Contact Source use + Create new attribute (or skip them), and do the same for Contactability Score (a Number attribute lets you filter on it). To skip a column, leave it unmapped.
  3. On Review values, check for Invalid format errors. Attio skips a value you don't fix, and "If all attributes for a record are invalid or blank, the record won't be imported."
  4. On Preview import, read the summary of changes the import will make. With Record ID mapped, you're only updating existing companies. Then click Start import.
  5. On the progress page, each record shows a green check or a red error; hover an error to read it. "No records found" means a Record ID didn't match. Start with a 10-row slice and check those companies before importing the rest.

Limits from the same page: CSV only, header row required, up to 100,000 rows, 100 columns and 100 MB. Attio's troubleshooting page says to run one import at a time. We found no documented undo for an update import, so keep your original export as the backup.

Then filter Companies on Contactability Score ≥ 50 and Email Deliverability READY or PARTIAL. That's your call and email list, sorted by how reachable each company is.

What Attio's docs don't say (so we don't claim it)

Sources and what we actually ran

Claim Status
prepare and merge on the made-up 8-row Companies export + hand-written EXAMPLE dataset (output pasted above) RAN, 2026-11-19
Export view as CSV / Excel; visible columns become file columns; Save for everyone for filters; enriched data exported blank; export restrictions; formula-injection setting READ, Attio Help Export lists and views, 2026-11-19
Unique attributes for Companies (Domains or Record ID); unmatched rows create new records except with Record ID; Record ID column in every export; multi-value updates are additive READ, Attio Help Bulk update records or lists via CSV import, 2026-11-19
Empty values skipped; only mapped attributes updated; Record ID unmatched rows skipped; Import / Export → Import CSV; mapping, review, preview; 100,000 rows / 100 columns / 100 MB READ, Attio Help Import data into Attio via CSV, 2026-11-19
Phone numbers must be E.164 and are validated READ, Attio Help CSV import formatting guide, 2026-11-19
"No records found" error; one import at a time; duplicate cleanup READ, Attio Help Troubleshooting CSV imports, 2026-11-19
Companies system attributes (no phone/email); Name is enriched READ, Attio Help Manage standard objects, 2026-11-19
Phone Number attribute type; "Email" object attribute on companies READ, Attio Help Create and manage attributes, 2026-11-19
No custom email-address-type attributes; Record ID is a UUID READ, Attio developer docs Email address, Standard objects, 2026-11-19
Actor pricing ($0.003 per website with contacts + $0.00005 start) READ, public Apify Store API, 2026-11-19
Attio export or import in a real workspace; live Actor run DID NOT RUN, $0 spent

Last tested: 2026-11-19. Corrections welcome via the product order page linked from Shop.