Google Ad Transparency Scraper

Collect public creative records from Google's Ads Transparency Center. Search by advertiser ID, advertiser name, or domain and get structured records with creative identifiers, advertiser details, formats, display dates, and available media links. The Actor uses the public Transparency Center responses and does not require a Google login.

Returns: one row per ad creative with creativeId, advertiserId, advertiserName, format, firstShown, lastShown, daysActive, imageUrl, previewUrl, sourceUrl. Input: {"advertiserIds": ["AR16735076323512287233"], "region": "US", "maxItems": 20} (also domains like nike.com, or searchTerms advertiser names) Best for: competitor ad monitoring, ad creative research, advertiser lead lists.

{"id": "CR11238584882709921793", "creativeId": "CR11238584882709921793", "advertiserId": "AR16735076323512287233", "advertiserName": "Nike, Inc.", "format": "text", "firstShown": "2023-11-16T22:30:36.000Z", "lastShown": "2026-10-01T19:22:17.000Z", "daysActive": 1051, "sourceUrl": "https://adstransparency.google.com/advertiser/AR16735076323512287233/creative/CR11238584882709921793?region=US"}

What does Google Ad Transparency Scraper do?

This Actor queries the public Google Ads Transparency Center for advertisements associated with an advertiser or domain. It turns the site's creative search response into one dataset row per creative. You can start with an AR advertiser ID, paste a Transparency Center advertiser URL, provide an advertiser name for suggestion matching, or search a domain. Multiple values can be combined in one run. A global maxItems limit caps the number of rows across all searches.

The Actor can filter by country and by creative format. The default region is the United States. Set region to anywhere when you want no regional restriction. The format filter accepts all, text, image, and video. Optional shownFrom and shownTo (YYYY-MM-DD) keep only ads whose display period overlaps that date range; they are applied after fetching, so narrow ranges scan more pages. If you enable details, the Actor makes an additional public request for each returned creative and includes numeric region codes supplied by Google. This takes longer than the basic search.

The source is Google's public Transparency Center, but this Actor is an independent tool. It is not affiliated with or endorsed by Google. The site's response structure can change. If it does, the Actor reports an error instead of inventing data. Results also depend on what the Transparency Center currently exposes for a particular advertiser, search term, domain, region, and format.

What data can you get?

Each dataset row represents one creative. Fields that the public source does not supply are null. The overview view highlights the fields most useful for scanning a result set.

Field	Meaning
id	Stable creative ID used to deduplicate rows within the run.
creativeId	Creative ID returned by the public search response.
advertiserId	Advertiser ID returned for the creative.
advertiserName	Advertiser name shown in the search result.
format	Creative format: text, image, or video when identified.
firstShown	First display timestamp in ISO date format.
lastShown	Last display timestamp in ISO date format.
daysActive	Inclusive days between firstShown and lastShown; null if either is missing.
imageUrl	Archive image URL when an image tag is present in the response.
previewUrl	Creative preview URL when Google provides one.
domain	Domain used to find the record on a domain search.
regions	Numeric region codes from the optional detail request.
regionCodes	ISO country codes (for example US) mapped from regions when includeDetails is on.
sourceUrl	Transparency Center page for the creative.

The image URL and preview URL are references to media or preview content; this Actor does not download or render the asset. A text format creative can still have an image URL because the source represents some rendered text ads as archived images. firstShown and lastShown are converted from the source's Unix timestamps. An active ad may have a recent lastShown value. The domain field identifies the search input that found a row and is not a claim that the advertiser owns that domain.

How to use

Open the Actor input form and enter at least one advertiser ID, advertiser name, or domain. For the most precise search, use an advertiser ID from a public Transparency Center advertiser page. If you use an advertiser name, the Actor requests Google's search suggestions and selects an exact case insensitive match when available; otherwise it uses the first advertiser suggestion. Review the returned advertiserId and advertiserName when a name could refer to several entities.

Choose a country code such as US if the geographic scope matters. Select anywhere to omit the regional filter. Choose a format or keep all. Set maxItems to the largest number of unique creative rows you want. The Actor paginates while a next page token exists and stops once it has pushed that many rows. If multiple searches find the same creative, the Actor emits it once.

Enable includeDetails when numeric region codes are useful. This makes a detail request for each selected creative and may expose an additional preview link. Detail failures are reported in the log; the search record is still returned. Apify datacenter proxy is enabled by default because Google rate limits direct requests. Blocked requests are retried with backoff and a fresh proxy session.

A search with no matching creatives, an unresolved name, or a target that stays blocked completes (SUCCEEDED) with an empty dataset and a status message saying what to change. An invalid region, missing target, or changed response structure produces a clear error. The Actor never puts error messages or placeholder rows in the dataset. For reproducible downstream processing, rely on the field names in the table and allow nullable values.

Input

Field	Input	Default and behavior
advertiserIds	Array of AR IDs or advertiser page URLs	Optional search target.
searchTerms	Array of advertiser names	Optional search target resolved through suggestions.
domains	Array of domain names	Optional search target for domain matching.
region	Two letter country code or anywhere	US.
format	all, text, image, or video	all.
maxItems	Integer from 1 to 10000	50 unique rows across the run.
shownFrom	Date YYYY-MM-DD	Optional; skip ads that stopped running before this date.
shownTo	Date YYYY-MM-DD	Optional; skip ads first shown after this date.
includeDetails	Boolean	false; request per creative details when true.
proxyConfiguration	Apify Proxy settings	Apify Proxy enabled by default; keep it on.

Supply at least one value in the first three arrays. The form shows example values as prefills; API callers should send their own targets. A page URL in advertiserIds must contain an AR identifier. Invalid identifiers are warned about and skipped. The region field accepts supported country codes from the Actor's region table. If a requested code is unsupported, use anywhere or choose another supported code.

{
  "advertiserIds": ["AR16735076323512287233"],
  "region": "US",
  "format": "all",
  "maxItems": 20,
  "includeDetails": false,
  "proxyConfiguration": {"useApifyProxy": true}
}

For a domain search, send {"domains":["nike.com"],"region":"US","format":"text","maxItems":40}. For a name search, send {"searchTerms":["Nike"],"maxItems":20}. Name matching can be ambiguous, so the returned advertiser fields should be checked before treating a run as belonging to a specific company.

Output

The example below is based on the Architect's live probe of the public creative search response. The exact set of live ads and display dates changes over time. Nullable fields are shown explicitly so you can see the record shape used by the dataset.

{
  "id": "CR11238584882709921793",
  "creativeId": "CR11238584882709921793",
  "advertiserId": "AR16735076323512287233",
  "advertiserName": "Nike, Inc.",
  "format": "text",
  "firstShown": "2023-11-16T22:30:36.000Z",
  "lastShown": "2026-10-01T19:22:17.000Z",
  "daysActive": 1051,
  "imageUrl": "https://tpc.googlesyndication.com/archive/simgad/10395020757433762867",
  "previewUrl": null,
  "domain": null,
  "regions": null,
  "regionCodes": null,
  "sourceUrl": "https://adstransparency.google.com/advertiser/AR16735076323512287233/creative/CR11238584882709921793?region=US"
}

When includeDetails is enabled, regions can contain numeric country identifiers returned by the detail service. A preview URL can be present for creatives that have a hosted preview. Different formats expose different media references. The Actor preserves the creative and advertiser IDs as strings so their leading digits and full precision remain intact.

How much does it cost?

This Actor is charged per result pushed to the dataset. One unique creative row is one result event. A duplicate creative found on a later page or in another search does not produce a second row. maxItems sets the maximum number of rows and therefore the maximum number of result events in a run. See the Actor's Pricing tab for the current terms. Proxy usage and other platform resources, if selected, are governed by your Apify account settings.

Use with AI agents (MCP)

Apify's MCP server can expose this Actor as a tool. Add the server URL below to an MCP client that supports remote servers, then enable the Actor tool. The exact client configuration interface varies by client.

{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com/?tools=axiomworks/google-ads-transparency-scraper"
    }
  }
}

Example prompts include “Find recent public creatives for advertiser ID AR16735076323512287233 in the US and return ten records,” “Search ads associated with nike.com and show the text format creative IDs and display dates,” and “Get public creative previews for this advertiser and include available region codes.” An agent should provide a target in its tool call because form prefills do not populate API calls. It should also read nullable fields as missing source data, not as proof that a media asset or region does not exist.

The structured dataset makes it possible to compare creatives across advertiser searches, organize research by last shown date, and save a set of source page URLs for review. A downstream agent can cite sourceUrl for a row. The MCP tool returns public records from the Actor's run; it cannot access private account data or ads hidden behind a login.

Integrations and API

Use Apify datasets to connect results to your own analysis. The examples below start the Actor with an explicit input and then read dataset items. Set the Apify API token in your environment or SDK configuration using your normal private setup. Do not place a token in a shared document or repository.

Python using the Apify client:

from apify_client import ApifyClient
import os
client = ApifyClient(os.environ["APIFY_TOKEN"])
run = client.actor("axiomworks/google-ads-transparency-scraper").call(run_input={
    "advertiserIds": ["AR16735076323512287233"], "maxItems": 10
})
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["creativeId"], item["sourceUrl"])

JavaScript using the Apify client:

import { ApifyClient } from 'apify-client';
const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('axiomworks/google-ads-transparency-scraper').call({
  advertiserIds: ['AR16735076323512287233'], maxItems: 10,
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
for (const item of items) process.stdout.write(`${item.creativeId} ${item.sourceUrl}\n`);

cURL to start a run uses the same JSON input:

curl -X POST 'https://api.apify.com/v2/acts/axiomworks~google-ads-transparency-scraper/runs' \
  -H "Authorization: Bearer $APIFY_TOKEN" \
  -H 'Content-Type: application/json' \
  -d '{"advertiserIds":["AR16735076323512287233"],"maxItems":10}'

The run response contains a dataset identifier. Retrieve its items through the Apify dataset API after the run finishes. If you need a fixed snapshot for analysis, save the dataset associated with that run rather than rerunning a live search and expecting the same advertisements or dates.

FAQ

How are duplicate ads handled? The Actor uses the creative ID as the stable row ID and emits that ID once per run. Duplicates across pages and search inputs are skipped. The first search that finds an ID determines its domain field.

Can I pass a public advertiser page URL? Yes. The Actor extracts its AR identifier and searches that advertiser. Other URL content is not fetched or used as an alternate source.

Why is a media URL null? The search response does not always include an archived image or hosted preview. Media references depend on creative type and the fields Google exposes for the record. The Actor does not guess a URL.

Does a name search always find the intended advertiser? No. Google can return multiple suggestions. The Actor favors an exact case insensitive name match and otherwise uses the first advertiser suggestion. Search by ID when precision matters.

What happens when the site blocks a request? The Actor retries blocked or rate limited responses with backoff. When Apify Proxy is selected, retry attempts use fresh sessions. A target that cannot recover ends the run with zero rows and a status message, so the dataset is not silently filled with fabricated records.

Why did a search return no rows? The target may have no visible creatives under the chosen region and format, the domain may not match a creative, or an advertiser name may not resolve. Try an advertiser ID, use anywhere, or select all formats and review the warning log.

Is it legal to scrape Google?

The Actor accesses public Transparency Center data without a login. Legal rules and site terms vary by location and use case. Evaluate your intended collection, storage, and redistribution against applicable requirements. Be considerate of the source and use modest request volumes. This Actor is independent of Google and provides source URLs so you can inspect the original public records.

Feedback

If a field looks wrong, include the run ID, the sourceUrl, the input used, and the affected field in a report through the Actor's Apify Store feedback channel. That gives a maintainer enough context to compare the dataset row with the live public response. Please avoid putting private account information or credentials in feedback. A change to Google's internal response format may require a parser update; a clear example helps locate it.