Google Ads Transparency Scraper: Ads and Creatives: facts

I built this Apify Actor. Store page: https://apify.com/garje/google-ads-transparency-scraper

FAQ

What does the Google Ads Transparency Scraper do?

It exports ads from the Google Ads Transparency Center by website domain, advertiser ID or advertiser name. Each ad is one row with its metadata and links to its content.

What input does it need?

At least one website domain, advertiser ID or advertiser name. You can add a region, a date window, formats, platforms and exact limits per advertiser and per domain.

What fields does each ad have?

Advertiser ID and name, creative ID, format, first and last shown dates, days shown, target domain and region. Content fields hold the headline and text where available, image and video links, the display URL and a contentStatus.

How much does it cost?

It costs $1.50 per 1,000 ads and $3 per 1,000 downloaded images. You pay only for rows and files saved, never for errors, retries or duplicates.

Why do most text ads come back as images?

Google publishes most text ads in the Transparency Center as one rendered image, not as separate words. In a measured run, 2,866 of 2,958 text ads (96.9%) came back as images, and contentStatus says imageOnly for them.

How do I monitor only new ads?

Set onlyNewSince to the previous run's time and schedule the task. Each run then returns only ads first shown after that time.

What if an advertiser name matches several advertisers?

Every advertiser Google suggests is scored and saved in the SEARCH_MATCHES record. By default only exact name matches, ignoring legal forms such as Inc. or GmbH, are scraped.

Which dates does it return?

firstShown, lastShown and daysShown came back for every ad in the regions we measured, US and anywhere. firstShownDate and lastShownDate are the dates as the Transparency Center displays them, in US Pacific time.

Can it download the ad images?

Yes, turn on downloadAssets and it saves up to 3 images per ad to the run storage. Each saved image is charged as a separate event.

What does it not do?

It does not read text out of images. It does not return the per-region first-shown dates and impression ranges that the Transparency Center shows only for EU countries.

What it does

Exports ads from the Google Ads Transparency Center by website domain, advertiser ID or advertiser name, with the ad content where Google provides it.

Inputs

• domains, advertiserIds or searchTerms (advertiser names): at least one is required
• region: a two-letter country code or anywhere
• dateFrom, dateTo: the shown window
• formats, platforms
• maxAdsPerAdvertiser, maxAdsPerDomain: exact limits
• onlyNewSince: only ads first shown after a time
• downloadAssets: save up to 3 images per ad
• minMatchScore, maxAdvertisersPerTerm: how advertiser names are matched

Outputs

One row per ad: advertiserId, advertiserName, creativeId, adUrl, format, firstShown, lastShown, daysShown, firstShownDate, lastShownDate, targetDomain, region, platform, headline, description, texts, displayUrl, imageUrl, imageUrls, videoIds, videoUrls, contentStatus, assetKeys, searchTerm, matchScore.

Pricing events

• ad: $1.50 per 1,000 ads saved.
• asset: $3 per 1,000 images saved, only when downloadAssets is on.

Limits

• Most text ads come back as images: in a measured run, 2,866 of 2,958 text ads (96.9%) were published as one rendered image.
• Limits per advertiser and per domain are enforced exactly.

What it does not do

• It does not read text out of images.
• It does not return the per-region first-shown dates and impression ranges the Transparency Center shows for EU countries.
• It does not need or use a login.

Changelog